Closely related: Does the NPVIC only take effect once enough states sign for it to have legal force? . That question asks whether, this question asks why.

The National Popular Vote Interstate Compact is written so that:

the agreement would go into effect among participating states only after they collectively represent an absolute majority of votes (currently at least 270) in the Electoral College (ref)

Suppose that the voters of one state, let's say Alaska, have decided that they're tired of waiting and they're going to throw their electors at the winner of the popular vote right now, regardless of the actions of any other state.

As I see it (and maybe I'm wrong) there are only possible two outcomes here. Either:

  1. The Electoral race comes in so neck-and-neck that Alaska's electoral votes are enough to swing the election. Alaska's three electors are enough to decide the election in favor of the national popular vote winner. The citizens of the state of Alaska get exactly what they wanted, a national popular vote! or...
  2. The Electoral race isn't close enough for Alaska's votes to matter. Ho-hum, the outcome is the same whether Alaska votes for the NPV winner or Mickey Mouse.

It seems like if a state wants to implement the NPVIC, there's no reason to wait for 270 votes. Yet clearly it was written that way for a reason. What am I missing?

Edit: David Hammen has mentioned in an answer:

If the laws of a state dictate that the state choose electors in accord with the nationwide winner of the election (but potentially against the will of the voters in that state), that would be perfectly valid. However, the laws enabling that elector selection would have to have been passed and enacted prior to the election.

For clarity's sake, that is the situation I'm interested in. A state dutifully, properly, and with ample time prior to any election, changes its laws so that it will, in future elections, award its votes to the NPV winner, without an interstate compact.

  • 3
    It is a bad idea to edit a question that changes the apparent intent of the question after answers have been given. Commented Mar 16 at 17:08

4 Answers 4


Consider the following scenario:

  • There are two parties, the dog party and the cat party
  • The state of EastWestNewArcaBama did sign the NPVIC, but is also a decidedly dog state with over 80% of voters voting for the presidential candidate of the dog party
  • The cat candidate wins the popular vote, but the dog candidate won the majority of the electoral college

Now if EastWestNewArcaBama were to follow the interstate compact and nominate the electors of the cat candidate, they would act directly against the interests of their own voters. 80% of them would get a cat as a president even though they wanted a dog. And they could have had a dog, if the state leadership hadn't decided to follow the NPVIC.

Well, that's the intention. But what if the situation had been reversed? With the dog winning the popular vote and the cat winning the electoral college? Then they couldn't have relied on a cat-state to follow the NPVIC and put their dog candidate into power.

The problem here is that if some states decide to unilaterally follow the idea of popular vote while others do not, then they weaken the political influence of their voters compared to those of the non-participating states. Which can be really hard to sell to the voters.

So the only way to sell to the voters "we changed the system, but in a way that is still fair to everyone" is when enough states are on board so that the new system works as intended in every case.

  • Now what happens if some states pass laws that say that they will only publicly report the number of votes received by the runner up before January 21 of the following year if the runner-up candidate requests such release, and the Dog party candidate wins that state, and almost wins a majority of the votes elsewhere, but the Cat party candidate doesn't want the winning margin reported? The state's only obligation under the Constitution would be to award its Electoral votes to someone, and no other states would have standing to know the winning margin.
    – supercat
    Commented Mar 15 at 14:52
  • 3
    @supercat I think Congress needs to know that the electors were awarded legitimately so it can properly certify them on Jan 6.
    – Barmar
    Commented Mar 15 at 14:59
  • @Barmar: If the runner-up candidate publicly concedes the the state, why would that not suffice?
    – supercat
    Commented Mar 15 at 15:23
  • Comments have been moved to chat; please do not continue the discussion here. Before posting a comment below this one, please review the purposes of comments. Comments that do not request clarification or suggest improvements usually belong as an answer, on Politics Meta, or in Politics Chat. Comments continuing discussion may be removed.
    – CDJB
    Commented Mar 18 at 10:03

Having the Compact take effect only when enough states cause it to "win" simplifies things -- the NPVIC either completely eliminates the electoral vote system, or else it does nothing. No ones needs to worry or even hear about what happens if only X electoral votes worth of states pass it. There's no scary analysis of every possible combination of states adopting it and how some might be bad. We don't have to think about how election night looks with some states having adopted and some not. It eliminates that category of legal challenges.

And there's no benefit to having states do it without a majority. The point of the compact is to have a simple "most-votes-for-president-is-the-winner" election, and make it so no one ever has to hear about electoral votes again. Without 270 worth of signees, we don't get that. The vote-loser can still be the next president, and we still have to add-up electoral votes, complicated even more with this new thing some states are doing.

  • The other answers go into all kinds of details, but I think this simple idea is what was really behind it.
    – Barmar
    Commented Mar 15 at 21:05

The states that voted the NPVIC believe that the desirability of possible election results is as follows, from most to least desirable:

  1. The state's favourite candidate gets to be president because he won the popular vote.
  2. The state's favourite candidate gets to be president despite not winning the popular vote.
  3. Another candidate gets to be president because he won the popular vote.
  4. Another candidate gets to be president despite not winning the popular vote.

Right now, all four options are possible. When the NPVIC triggers, only options 1 and 3 will still be possible. Option 2 changes into option 3, which is a bad thing, but option 4 changes into option 1, which is a good thing. It's a net positive, if you consider both to be roughly equally likely.

When a state would do what you propose, option 2 changes to option 3, but option 4 doesn't change to option 1. That's clearly a net negative.

Your theory would be right if some state believed that options 2 and 3 are reversed. So they want to see the popular vote honoured, even if it means the "wrong" candidate gets to be president. I've never seen any evidence of a state believing so. States want to see their preferred candidate in office if they win the popular vote. The fact that another candidate might get into office this way is an acceptable tradeoff, but it's not at all the goal.

  • Right now, in 1. and 3. it should not be "because", but "and". Commented Mar 16 at 0:21
  • I accepted Owen Reynolds' answer because it most clearly answered the question I asked, but I want to highlight your answer as being the one that most clearly determined where I was mistaken. Thank you!
    – Steve V.
    Commented Mar 23 at 23:39

To answer the OP's question, consider some of the "flaws" in the plan as pointed out by https://www.cato.org/blog/fatally-flawed-national-popular-vote-plan.

  1. A state that doesn't like the plan could (as North Dakota's senate did) pass a law that has the state withhold its popular vote totals until after the Electoral College votes in December. [This provision was not enacted into law; the North Dakota's House removed it].

  2. A state could (as Alabama did in 1960) refrain from putting the name of the presidential candidates on the ballot, but instead have direct elections of the electoral college members from that state; who would not pledge their votes before the election to any candidate; but instead after they are chosen, they would deliberate and chose which presidential candidate to award Alabama's electoral college votes. Therefor it would not be possible to say what the popular vote was for any presidential candidate from that state.

  3. A state could (as Maine has done) implement ranked choice voting for president; where based on the voters indicated first, second, and third preferences, a calculation is done to find the person most favored by the state. Here again it is difficult to state in some scenarios who won the popular vote.

These (and other) issues were likely anticipated by the original authors of the bill; and they presumably thought that adding the clause that "the bill go into effect among participating states only after they collectively represent an absolute majority of votes (currently at least 270) in the Electoral College" would be sufficient to make such responses ineffective, as the majority of the states would already be implementing a direct presidential election.

  • 3
    I think these are still fatal flaws even if the majority is reached. But it might be politically unpopular for non-NPV states to take these actions.
    – Barmar
    Commented Mar 15 at 21:07
  • But it might in many cases be possible to determine who won the popular vote even with a few renegade states. But if all the states that don't join the compact take one of these actions, it could spoil it.
    – Barmar
    Commented Mar 15 at 21:09
  • 1
    @Barmar The thing is: If the NPVIC states already represent the majority of the electoral college, then they don't need to put up with such shenanigans. When a state doesn't or can't come up with a popular vote count, then the NPVIC states could just ignore the election results of that state completely and still elect a president based on the counts from the states that provide them. So the non-compliant states would either have to report a popular vote count or have no say in the presidential election whatsoever.
    – Philipp
    Commented Mar 15 at 21:32
  • @Philipp They could, but I don't think that's what the language of the NPVIC says. So they'd be changing the rules after everyone joined.
    – Barmar
    Commented Mar 15 at 21:36
  • 3
    @Barmar: The NPVIC folks explain this by saying (in the FAQ below the actual text) that they would exclude any election which does not list presidential candidates on the ballot, because it would not be a "statewide popular election" within the meaning of the agreement. They specifically point to Alabama in 1960 and claim it would not pose a problem for them. As for North Dakota, it would violate 3 USC 5(a)(2)(A) to do that, so other states probably will not choose to do so. I don't know what to tell you about Maine and RCV, though.
    – Kevin
    Commented Mar 16 at 8:17

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .