Electoral College Amendment

Draft amendment. Not yet rated or edited by the WeAmend community.

Draft EC-20-07

Electoral College Amendment

Section 1. The Federal election for President and Vice-President is direct and rank-ordered. Each voter shall indicate on the ballot their first, and may indicate their second and third rank-ordered choices from among the remaining candidates for President and Vice President. No candidate may appear on the ballot or in the rankings more than once.

Section 2. After election day, but no later than the Monday after the second Wednesday in December, the Secretary of State of each state and the District of Columbia will transmit their official state voting results directly to both houses of Congress. Votes not received by the end of that date are invalid. Congress may provide by law for the tallying and approval of the received vote results, except that the final tally and approval must be held by the newly elected Congress.

Section 3.  Among the votes accepted by Congress, if no candidate for President receives a majority of cast first-choice votes, the second-choice votes will be added to the first choice. If no candidate receives a majority of the sum of the first and second choice votes, then the third-choice votes are added and summed, and the candidate with a majority or plurality of summed votes is elected president. An identical procedure is used for the Vice-president. If any candidate for either office dies or otherwise withdraws from consideration before the final tally is ratified by Congress, their votes are treated as if they were still available to serve.

Section 4. This amendment shall be effective three years after ratification.

Has the Electoral College outlived its usefulness? And if so, what is our best alternative? The answer is an amendment to vote directly for President in a modified ranked-choice election called RCV123.

CLICK to open/close for details on wording of amendment

Section 1: Mandates ranked choice voting. The President and VP are chosen in separate elections, but in most cases are expected to be nominees from the same party. Currently, the VP choice is often based on filling a geographical, age or race check box- the public rarely has a say. It is possible voters may prefer one candidate for President, but highly dislike their vice-presidential choice (many states split the ticket1). VPs matter- fourteen VPs succeeded the president in office. The possibility of a split party ticket is an incentive for each candidate to treat their VP choice with respect.

Candidates must be listed only once on the ballot- multiple appearances would allow them to claim first, second and third rank votes, undermining RCV123. However, while their name can appear only once on the ballot, they are free to list multiple party endorsements.

Unless modified by a new voting rights amendment, current state and federal rules governing voting eligibility (such as residency requirements or voter identification) continue to apply.

Section 2: Requires states to complete their voting tallies in the same timeframe as the Electoral College- around six weeks. Congress can direct the states to submit detailed voting results, i.e. an anonymous rank order for each voter, or a state-wide total for every permutation of rank orders. As is true today, Congress will determine the process for validating, tallying and certifying the results.

In an alternate version of Section 2, Congress may adjust the tally date (to allow for unforeseen circumstances), but no later than two weeks before the Inauguration.

Section 3: Describes how ranked choice voting RCV123 is implemented procedurally. Note Congress has the option to “accept” the state tallies- on rare occasions, they may reject a state vote if there is strong evidence of malfeasance. Presumably following the same rules as the Electoral College (e.g. a majority of both houses of Congress must vote to reject).

In a compromise bow to state’s rights, the amendment could defer to state attorney generals to validate their results, and not involve Congress in this step.

In the rare case when a candidate dies or withdraws before the vote tally is ratified by Congress, the will of people prevails. That is, had the candidate died the day after ratification, the 25th amendment would apply. Presumably, the people chose the VP knowing they might have to replace the President on short notice while in office- the same logic applies during ratification2

Section 4 After ratification, at least three years elapses, allowing time for the states and Congress to prepare.

Under the Constitution, the President and Vice-President are confirmed indirectly through a system known colloquially as the “Electoral College”. Voters choose electors and electors choose the president. Originally designed to balance an amalgam of slave-state concerns, practical limits to information in an age of horseback communication, sovereign representation among the colonies and fears that an unruly mob might embrace a demagogue, today it lingers on as a distorted vestige from the past.

The college is mortally wounded. Instead of nominating wise leaders to cast their votes for the best candidate, members of the electoral college are no longer free to reassign their votes based on circumstances.  They are locked down by statute to a specific candidate.  Perhaps just as well, since few voters care about their elector’s abilities other than their party affiliation. In fact, because 48/50 states enforce “winner-take-all” rules, the President can lose the popular vote and dominate the college to win, as has occurred five times in history. With an additional half-dozen close calls. Undermining electoral legitimacy in the majority of American’s eyes.

The Electoral College creates the appearance of a mandate, but in fact squeezes out third-party candidates and decides winners based on the thinnest of margins3. In Bush v Gore, fraud and ineptitude enabled 537 disputed Floridian votes to decide an election for the entire country- despite a clear half-million popular vote advantage, and tens of thousands of uncounted Floridian ballots. Legal, but hardly a legitimate outcome.

Despite two-party control over primaries, third party candidates sometimes attract huge followings ( e.g. Ross Perot (19%) and George Wallace (14%)), winner-take-all statutes locks them out of meaningful gains in the Electoral College. Tipping the balance from a popular democracy to a rigged political game.

One can argue the Electoral College- an echo from debates over the proper workings of a Federal representative democracy- is functioning as intended by granting electoral college seats based on the size of the congressional delegation, not population. Tilting the playing field towards the states and away from direct popular vote. But this hypothetical value (even if desirable) is ephemeral- in fact, because of the relentless math of winner-take-all, large population battle-ground states dominate the political calculus. The intended small state advantage remains unfulfilled. In any case, equal representation in the Senate already provides significant voice for small states. The Electoral College does not.

There are no “perfect” election systems– any list of desirable criteria contains elements which are mutually exclusive4, and there will always be times when any electoral system fails to perform as hoped.  Or is too complicated for the average voter to navigate. More often than not, undermining universal participation and legitimacy.

In our view, a more democratic system would:

  • Assure that voters from every state have an equal voice and opportunity to elect a candidate of their choice.
  • Enable third party candidates to gain national traction and a shot at electoral success.
  • Incentivize candidates to adopt policies that benefit all Americans.
  • Incentivize candidates to campaign in both large and small states.
  • Encourage candidates to choose more highly qualified vice-presidents.
  • Ameliorate the effects of voting irregularities in one state corrupting the national outcome.
  • Adjusts smoothly if a candidate dies or withdraws before Congress ratifies the state vote tally.
  • Encourage voters to see the ballot box as part of the solution5.
  • Increase voter turnout.

How can we achieve these goals?

One might leave the Electoral College alone, but modify the rules for assigning electors. For example, eliminate winner-take-all by statute or amendment. Or decrease the size of congressional districts, which would increase the number of congresspersons and thus dilute the small state numerical advantage. Or assign electors to the winner of each district. Or increase the number of states. Or turn blue states purple by migration. These adjustments would more closely align the popular vote with the Electoral College outcome. Alternatively, clever (but highly suspect and fragile) schemes like the National Popular Vote Interstate Voting Compact force state electors to vote in synchrony with the national majority. Creating a popular direct vote by fiat.

Unfortunately, this still leaves party politics and state house lobbyists in charge of the election. Instead of indirect solutions, we choose this opportunity to bring a modified version of ranked choice voting to the national stage.

In a ranked choice voting system (RCV), you fill in the ballot with your preferred candidate, and then, in order of preference, their best alternates. If no candidate receives a majority on the first round, the last place candidate is removed from consideration, and their ranked alternatives are reassigned to their desired candidates. Rarely in the US do more than three candidates make it on the ballot with any chance of winning, but in principle a dozen or more could run.

RCV is used worldwide, though much less commonly than simple majority voting. RCV encourages third party candidates to participate, and in some circumstances, enables a consensus candidate to emerge after the initial round. RCV is particularly well suited for multi-member elections. But traditional RCV exhibits numerous flaws which must be addressed before inclusion in a constitutional amendment.

CLICK to open/close for more details on RCV flaws

These four scenarios compare majority voting w/wo the Electoral College, and conventional RCV. Under some circumstances they produce democratic results, but in other cases, fail, sometimes spectacularly.

Scenario I- Plurality and Electoral College that works. Three candidates run, one candidate receives 60% of the popular vote, 70% of the states and 80% of the Electoral College. They win either way. Majority rules, and the majority is satisfied.

Scenario II- Plurality and Electoral College voting that fails. Three candidates run, receiving 43%(A), 34%(B) and 25%(C) of the vote.

  1. The 43% “A” candidate wins in battle ground states, dominates the Electoral College, and becomes President. The majority of voters are not satisfied, and most people think the system has failed somehow to reflect “democratic” principles.
  2. No candidate receives a majority of Electoral College votes, despite winner-take-all. The contest is thrown to the House in a “Contingent” election, which might have very different politics and demographic representation than the general public.

Scenario III- Rank Choice voting that works. Three candidates run, receiving 43%(T), 34%(B) and 25%(S) of the vote.

  1. Like Biden/Sanders, the “B” and “S” candidates agree more than they disagree. In RCV the third-place candidate “S” is dropped, and their second-choice candidates receive their votes. Most of the “S” voters choose “B” as their alternative because they understand how RCV works (and even a few of the “T” voters list the “B” candidate as their second choice). The “B” candidate wins with 57% of the vote, and clearly satisfied most of the voters most of the time. RCV also encouraged the third party “S” candidate to run in the first place, which improved the quality of the debate.
  2. A parliamentary approach. B, S and T candidates will never agree. But, before election day, B and T strike a grand compromise on policy and cabinet positions, and B instructs his supporters to vote for T as their alternative (if B instructed them to simply vote for T on the first round, many would not bother to vote. A transferable second round vote reminds T of the strength of B’s base). T wins as a coalition candidate.

Scenario IV- Rank Choice voting that fails. Three candidates run, receiving 43%, 34% and 25% of the vote. In RCV the third-place candidate is dropped, and their second-choice candidates receive their votes. This can fail in multiple ways:

  1. The electorate is so polarized that very few second-choice votes are cast. So the 43% candidate wins, with less than enthusiastic majority support after all rounds. RCV is at best irrelevant, or amplifies the concerns of fringe voters. This low compliance issue becomes an even worse outcome if five candidates run, and the eventual winner started with 25% first round votes, and ends with 35%.
  2. Polarization and protest votes dominate the first round. The third-place candidate is widely admired, but only as the second-choice alternative. In fact, 80% of the electorate backs her as their alternate choice. But, under RCV she is dropped after the first round, although she is actually the consensus candidate.
  3. Without a national uniform ballot, one of the candidates only makes it on half the state ballots. Party politics and onerous qualification criteria keep them off. Even though polling indicates they are likely to take second place after the first round if they appeared on the ballot everywhere, they end up a distant 4th. RCV cannot remedy the lack of national ballot standards.
  4. A parliamentary approach. B, S, L, V and T candidates will never agree. But, before election day, B, L and S strike a grand compromise on policy and cabinet positions, and LS instruct their supporters to vote for B as their alternative. B wins as a coalition candidate, but the coalition quickly falls apart. Unlike a parliamentary government, this saddles our nation with a President without a mandate, damaging their ability to govern effectively for the next four years. And damaging the reputation of RCV.

Practically speaking, it’s very hard for humans to dependably rank-order beyond three candidates or weight each candidate by some logical percentage, despite the best hopes of policy wonks. Few voters are willing (or sometimes able) to research a half dozen candidates and feel confident they have made the “best” decision. High information barriers and complexity may depress voting rates, cause spoiled ballots, or introduce randomness into the count. In a polarized electorate, partisan voters can’t envision a second-choice alternative, let alone a third. More sophisticated voters might prefer to adjust their rankings if they knew, ahead of time, the first-round results. They might desire a run-off election over RCV.

And RCV can lead to perverse outcomes. In this highly realistic example, imagine polarization and protest votes dominate the first round of a three person race. The third-place candidate is widely admired, but only as the second-choice alternative. In fact, 80% of the electorate backs her as their alternate choice. But, under RCV she is dropped after the first round, although she is actually the consensus candidate.

For all of these reasons, a modified version of RCV is envisioned. We call this RCV1236. Unlike traditional RCV, you can only rank your top three candidates. This reduces cognitive load, and encourages a plethora of weak candidates to negotiate with the front runners to obtain their support as the election nears. In RCV123 the lowest ranked candidates are not dropped after each round. This avoids prematurely eliminating consensus candidates. After three rounds, if a majority candidate does not emerge, the plurality leader is selected. Which accurately reflects the split political opinions of the country (unlike the Electoral College, which tries to hide this division by rounding up states under winner-take-all). RCV123 is tolerant of first round protest votes, prevents consensus candidates from premature elimination, discourages highly partisan voting as ineffective, and leads to generally fair outcomes in a wide range of realistic election scenarios.

Although WeAmend argues Constitutional obedience should be reasserted by each generation, it is interesting to note that James Madison recognized these very same deficiencies in the Electoral College, and proposed a form of RCV. As the historian Alexander Keyssar remarks:

Madison proposed 7 a different mechanism: each elector would cast two ballots for president, one for his first-choice candidate and the other for his second choice. A candidate would become president if he received a majority of the votes on the “first choice” list or (if that did not occur) on the “second choice” list. “Such a process,” Madison wrote, “would avoid the inconvenience of a second resort to the Electors; and furnish a double chance of avoiding an eventual resort to Congress.”

Why Do We Still Have the Electoral College? Alexander Keyssar, Harvard University Press (July 31, 2020)

Simple. Clear. Fair. Democratic.

  1. Currently, 26 states elect a lieutenant governor on a ticket with the governor, while 17 states elect a lieutenant governor separately. In West Virginia, the President of the Senate, as elected by the State Senators, serves as the state’s lieutenant governor. In Tennessee, the State Senators elect a Speaker of the Senate, who in turn serves as lieutenant governor. Five states do not have a lieutenant governor.
  2. One argument to retain the Electoral College is to handle rare cases. For example, it after the election it became known the winning candidate was a spy working for an adversary. The Electoral College could over-rule the duped voters and select the VP or another candidate in their stead. But the EC is hardly a wise deliberative body with the respect necessary to upset a national election. In this case, our democracy would rely on the impeachment process or public pressure to oust the spy, deferring to the democratically elected VP. Similarly, if both candidates died in a freak accident, they would still win the election, but again the 25th amendment applies.
  3. Besides the infamous Hayes-Tilden scandal of 1878, Cleveland won the 1884 contest by taking all of NY State’s electors by a margin of 1,149 votes.
  4. PlusMaths offers a simple introduction to voting systems and their strengths and weaknesses
  5. As John Lewis testified to Congress in 1979, by eliminating Winner Take All it would be “much easier.. to register new people” in the South if they believed their vote would count in the final tally
  6. RCV123 is a variant of Bucklin Voting .
  7. From James Madison to George Hay, 23 August 1823