Every animation on the site, shown as still images at its key moments, with a fuller explanation of each step.
Figure 1. Votes that never reach the Electoral College
Two states from 2024, one won by each party, show what winner-take-all does to the voters on the losing side. See it in context.
Figure 1a. Wisconsin and New Jersey in 2024. Each dot stands for about one-fortieth of the votes for the two major candidates. Wisconsin’s dots split almost exactly in half, with Trump ahead by 0.9 points; New Jersey’s lean toward Harris, who won by 5.9 points. Below each state, its electors wait to be filled: 10 for Wisconsin, 14 for New Jersey.Figure 1b. Winner-take-all is applied. Every Wisconsin elector goes to Trump and every New Jersey elector to Harris, and the losing side’s dots turn gray where they sit. In Wisconsin, 1.67 million Harris voters end up with no electors; in New Jersey, 1.97 million Trump voters do. Nearly half of each state’s voters are left with no one representing them.
Figure 2. Electoral votes are a collection, not a weight
A state’s electoral votes are usually thought of as a single number. They’re really a group of individual electors, and nothing requires them to all go the same way. See it in context.
Figure 2a. Electoral votes as they’re usually pictured: one solid weight per state. Wyoming has 3, Wisconsin 10, Pennsylvania 19, and California 54, each drawn as a circle sized to its count.Figure 2b. Each weight is really a collection of separate electors. The solid circles dissolve to show them: 3 pebbles for Wyoming, 54 for California, one for each person who will cast an electoral vote.Figure 2c. Nothing requires a state’s electors to all go the same way. Colored as Fair Slate would assign them from the 2024 vote, each state splits to match its voters: Wyoming 2 Trump and 1 Harris, Wisconsin 5 and 5, Pennsylvania 10 Trump and 9 Harris, and California 32 Harris, 21 Trump, and 1 for Kennedy.
Figure 3. Three balanced compacts on the map
Pairs of states that lean opposite ways can switch to Fair Slate together, so neither party loses ground. See it in context.
Figure 3a. The 2024 electoral map, one square per state, colored by who won it. Six states are outlined: three pairs that could form compacts, each pairing a state Trump won with one Harris won. Missouri pairs with Colorado, Tennessee with Maryland, and Montana with New Hampshire.Figure 3b. The six states lift off the map into three panels, so each pair can be seen side by side. Under winner-take-all, every state’s electors are a single color. Together these six states send 25 electors to Trump and 24 to Harris.Figure 3c. Each pair switches to Fair Slate at the same time. Every state now sends electors from both sides. Because each pair is balanced, the electors that change hands cancel out: the six states still total 25 for Trump and 24 for Harris, but now both sides’ voters win electors in every state.
Figure 4. The spoiler effect under winner-take-all
When a state awards every elector to the top candidate, a few votes shifting to a third candidate can flip the whole state. See it in context.
Figure 4a. A state’s votes on a balance scale. Candidate A’s side holds 18 votes and Candidate B’s holds 17. A has one more vote, so under winner-take-all, A takes every elector.Figure 4b. Two of A’s voters choose a third candidate, C, instead. A drops to 16, the scale tips, and B now takes every elector. Two votes moved the entire state’s electors from one candidate to another, even though C won nothing. That is the spoiler effect.
Figure 5. The same spoiler, with Fair Slate
The same 35 votes as the scale above, divided among 5 electors in proportion to the vote. See it in context.
Figure 5a. The same 35 votes as before, 18 for A and 17 for B, with 5 electors to fill.Figure 5b. With Fair Slate, each elector stands for 7 votes. A’s 18 votes fill three electors and B’s 17 fill two, so A gets 3 and B gets 2.Figure 5c. Two A voters choose C instead. A’s votes now fill only two electors, so one elector crosses over to B: A 2, B 3, C 0. The same two votes that flipped all 5 electors under winner-take-all move just one here.
Figure 6. A close count in Florida, 2000
Bush won Florida by 537 certified votes out of nearly 6 million. Both grids hold the same votes; only the rule for awarding electors differs. See it in context.
Figure 6a. Florida’s 25 electors, drawn twice. Each elector holds four pebbles, and each pebble is about 1% of the vote. Bush and Gore each fill 12 electors; the elector in the center goes to Nader, holding his 2 pebbles plus one leftover each from Bush and Gore. The top-right pebble of that center elector is the swing vote: the margin. Here it’s counted for Bush. Under winner-take-all, on the left, all 25 electors are Bush’s.Figure 6b. The swing vote flips to Gore. Under winner-take-all, every elector flips with it, and all 25 go to Gore. Under Fair Slate, on the right, nothing else changes: the margin sat in the leftovers that decided Nader’s elector, so it moved no electors at all.
Figure 7. How compacts hold up when a state leaves
The National Popular Vote compact works only above a threshold. Fair Slate compacts each stand on their own. See it in context.
Figure 7a. The National Popular Vote compact as one bar of member states, each block sized by its electors. Its members hold 222 electors today, short of the 270 marked by the line. Until it reaches 270, the compact does nothing.Figure 7b. If a few more states joined, the bar would cross 270, at 272 electors, and the compact would take effect. Gold marks it as active.Figure 7c. If one of those states later left, the bar would fall back to 257, below 270, and the whole compact would go dormant for every member at once.Figure 7d. Fair Slate compacts work differently. Each is a separate agreement among a few states, shown here as small groups of state tiles. If one compact ends, only its members return to winner-take-all; every other compact keeps working.
Figure 8. How the electors are chosen
Each party ranks its possible electors. The vote decides how many each party gets, and they’re seated from the top of each list. See it in context.
Figure 8a. A state votes 55% for Party A, 43% for Party B, and 2% for Party C, with 10 electors to fill. Ten tokens appear beneath the vote bar, each colored by the share above it. The sixth sits on the line between A and B and goes to A, which has the larger leftover fraction. Below, each party’s slate of possible electors is ranked from 1 to 7.Figure 8b. The tokens drop onto each slate from the top. A’s six go to its electors ranked 1 through 6, and B’s four to its top four. Party C’s 2% is not enough for an elector. The chosen electors take their party’s color; everyone else stays gray.Figure 8c. The chosen electors move into a single column, and the state certifies them as its slate, shown by the stamp.Figure 8d. The certified electors cast the state’s votes for president, and Congress counts them.
Figure 9. Why districts can be gerrymandered
Awarding electors by district, as Maine and Nebraska do, lets the people who draw the lines shape the result. See it in context.
Figure 9a. A state of 50 voters, 26 Republican and 24 Democratic, with 7 electors: one for each of 5 districts, plus 2 for the statewide winner. With the districts drawn as horizontal rows, Democrats win 4 of them, so the electors go 3 Republican and 4 Democratic, a majority for the side with fewer votes. On the right, Fair Slate splits the same voters 4 to 3, with no lines to draw.Figure 9b. The same voters, with the lines redrawn as vertical columns. Not one vote has changed, but now Republicans win 4 districts, and the electors go 6 to 1. Fair Slate still gives 4 to 3.
Figure 10. Why a larger House makes the result more precise
Electors are whole people, so a state with few of them has to round coarsely. Adding House seats gives every state finer steps. See it in context.
Figure 10a. A small state with 36 votes, 24 Republican and 12 Democratic, and 4 electors: two for its senators and two for its House seats. The bar at the top shows the vote, 67% to 33%.Figure 10b. Each of the 4 electors has to hold 9 votes. Three Democratic votes end up in a Republican elector, and the split rounds to 3 to 1. The second bar shows the electors at 75% to 25%, 8 points off the vote.Figure 10c. Add two House seats, for 6 electors of 6 votes each. The votes redistribute, those three Democratic votes join an elector of their own, and the split is 4 to 2. The electors match the vote exactly, 67% to 33%.
Figure 11. Largest remainder
A toy state with 5 electors and 100 votes: A 49, B 32, C 19. See it in context.
Figure 11a. Each candidate’s share of the 5 electors is its share of the vote times 5: A 2.45, B 1.60, C 0.95.Figure 11b. Each candidate first gets the whole-number part of its share: A 2, B 1, C 0. That fills 3 of the 5 electors.Figure 11c. The two leftover electors go to the largest fractions: C’s .95, then B’s .60. A’s .45 misses. Result: A 2, B 2, C 1. C wins an elector with less than one elector’s worth of votes.
Figure 12. Threshold
The same votes, with candidates below one elector’s worth dropped first. See it in context.
Figure 12a. One elector’s worth is 20 votes, marked by the dashed line. C has 19, just short, so C is dropped.Figure 12b. A and B split the 5 electors by their own votes: A 3.02, B 1.98. The whole parts give A 3 and B 1.Figure 12c. The leftover elector goes to the larger fraction, B’s .98. Result: A 3, B 2, C 0. C’s 19 voters elect no one.
Figure 13. Threshold with ranked ballots
The same cutoff, but C’s voters ranked a second choice, B. See it in context.
Figure 13a. C has 19 votes, short of 20, so C is dropped. Instead of being discarded, its votes move to its voters’ next choice, B, whose bar grows from 32 to 51.Figure 13b. Now A has 49 and B has 51: shares of 2.45 and 2.55, so 2 electors each.Figure 13c. The leftover elector goes to the larger fraction, B’s .55 over A’s .45. Result: A 2, B 3, C 0. C’s voters won no elector of their own, but they decided who got the last one.
Figure 14. D’Hondt, or Jefferson’s method
A divisor method: each candidate’s votes are divided by 1, 2, 3, and so on, and electors go to the highest results one at a time. See it in context.
Figure 14a. Each candidate’s votes divided by 1, 2, and 3. The highest number in the table, A’s 49, wins the first elector.Figure 14b. The next highest numbers win the next electors: B’s 32, then A’s 24.5. Each elector A wins means A’s next number is smaller.Figure 14c. C’s 19 wins the fourth elector, and A’s 16.3 takes the last. B’s 16, outlined, just misses. Result: A 3, B 1, C 1.
Figure 15. Electors against population share, by state
Every state’s electors compared with the number its population would earn, on a logarithmic scale. This chart doesn’t move; the image below is the whole figure. See it in context.
Figure 15a. One row per state, from the most over-represented at the top to the least at the bottom. Gray blocks are a state’s electors, one block per elector; the line is its population share, the number of electors its population would earn. Blocks beyond the line are extra electors, colored by which way the state has leaned over the last five elections; a dashed outline past the blocks marks electors a state is missing. At the top, Wyoming has 3 electors against a population share of 0.9, and Vermont, DC, Alaska, and North Dakota are similar. At the bottom, California has 54 against about 64, and Texas, New York, and Florida also come up short. The column on the right counts each state’s extra or missing electors: Wyoming has 2.1 extra, while California is missing 10.2 and Texas 7.3. A check marks the rare near-exact fits: Indiana, Maryland, and Missouri. Because the scale is logarithmic, big states’ shortfalls look small in the bars; the counts show they are the largest in absolute terms. The extra electors split nearly evenly by party: about 22 in Republican-leaning states and about 19 in Democratic-leaning ones.
