Blog

One of the main criticisms of the National Popular Vote interstate compact (NPV) is that in order to function properly it requires every state, including those not in the compact, to cooperate. If just one or two states don’t, it gets very messy—as in, epic constitutional crisis messy.

For non-member states, cooperation doesn’t necessarily mean actively working with member states to provide the vote totals needed to make the compact work. Simply producing and publicly releasing before the deadline either the state’s official canvas or Certificate of Ascertainment is all the cooperation needed for the compact to operate. So long as every non-member state does that, there shouldn’t be any problem as far as obtaining the necessary vote totals by the deadline.

So, how is that going in 2024?

Well, if the compact was in effect this time around, there would be a major, California-sized problem as well as a smaller, West Virginia-sized problem. In a closer election, such issues could affect the final result and thus have serious legal and political ramifications.

Let’s talk first about the deadline. Wednesday, December 11, was the “drop dead” date for every state to submit its Certificate of Ascertainment to the National Archives. This is the document that confirms which slate of presidential electors has been elected. It also includes vote totals for the electors. In past elections, December 11 would have been called the “safe harbor” deadline, meaning it was a soft deadline—recommended but not essential or required. After the Electoral Count Reform Act passed in 2022, it became a hard deadline—any state that doesn’t submit its own Certificate of Ascertainment by the deadline is at risk of not having its electoral votes counted by Congress.

For NPV’s purposes, this makes it vital that every state complete and certify its vote count and make it publicly available prior to the deadline, so that member states have time to obtain the vote counts of every other state, add them to their own, and then determine—using only official, certified vote totals—which candidate received the most popular votes nationwide and is therefore the winner. 

Back to California and West Virginia. As of Friday morning, December 13—two days after the deadline—neither state has certified its final vote totals, nor has either state made public its Certificate of Ascertainment (it isn’t clear if they’ve submitted them to the Archives either, but it is certain the Archives hasn’t received, processed, and posted either of them yet). Both have unofficial results available on their websites (here is a screen grab of the California site, and here is a screen grab of West Virginia’s), but it’s not clear if those unofficial results will be the final, accurate totals – it’s not unusual for thousands or even tens of thousands of votes to be added at the last minute before certification.

So, if NPV was in effect in 2024, what would happen if some states hadn’t completed their certification process and made the results publicly available before the deadline? Well, as a pair of lobbyists from the NPV organization testified in North Dakota a few years ago, the chief election official of each member state—usually the secretary of state—is still responsible for determining how many votes were cast in a state for which there are no official, certified vote totals available. 

But there’s no guidance or guardrails in the compact on how to make that determination, meaning they have free reign on the method used to estimate the vote totals to assign to each candidate. One obvious option would be to just use the unofficial vote totals that are available. Another would be to try to figure out how many votes are likely to be added to the unofficial totals before certification and then use that adjusted figure (and of course there are a variety of methods that can be used to make that estimation). 

Really, there’s no limit in the compact to the methods that might be tried. The lobbyists in North Dakota suggested methods that, if used in California or another populous state, could literally add millions of votes to the totals. 

And because the compact lacks a coordinating body or dispute resolution mechanism, there’s nothing to prevent member states from using different methods that produce different estimates and potentially even different winners. The chief election official in compact-member Massachusetts might, for example, opt to go with the unofficial results given that’s what is available by the deadline, while fellow compact-member Illinois might believe there are tens of thousands of votes likely to be added to the unofficial totals before certification and instead use estimated totals based on that belief. 

A fair response might be that California is a member of the compact and so it would certainly make sure to conclude its counting and certify the results ahead of the deadline. And, maybe? Probably? 

But that still leaves non-member West Virginia and its 763,390 (unofficial) votes split roughly 70/28 between Trump and Harris, and the choices for determining the number of votes to include from the Mountain State that would be available to the chief election officials of member states.

Another response might be that the difference between the projected, actual, and estimated vote totals are likely to be small and could not possibly affect the outcome, at least assuming a good-faith determination of the number of votes to assign to each candidate is used. For example, if a member state decided to apply West Virginia’s percentage of the vote reported for each candidate in the unofficial results to total known turnout—one of the several options described by NPV’s lobbyists in North Dakota—it would add roughly 2,300 votes to Kamala Harris’ total and about 5,700 votes to Trump’s, for a net gain of 3,400 votes for Trump in the national count.

But if this had to be done in a larger state such as California, the gains could be considerably more. As best I can tell from the available information, there are is a gap of roughly 277,000 ballots in California between turnout and the unofficial vote totals of the six candidates on the ballot. Applying the same methodology used on West Virginia’s vote totals to California’s numbers would give Harris a net gain of about 56,000 votes—not enough in 2024 to affect the national outcome, of course, but in a closer national election and combined with a few problems in other states, well, it’s not hard to see how this whole thing blows up into a crisis.

NPV is extremely vulnerable to this and other problems when it comes to actually adding up the votes, in particular when it comes to non-member states not following the timeline required by a compact they are not a party to. Even worse, the compact provides no guidance on how to resolve this issue, instead leaving each member state on their own to deal with it as they see fit. It’s a recipe for disaster even if everyone is acting in good faith, and I don’t even want to think about what happens if you have just a few bad-faith actors.