Blog

A critical weakness of the National Popular Vote interstate compact (NPV) is the lack of an official national vote count. Instead, the compact gives the chief election official in each member state—usually the Secretary of State—the power to “determine the number of votes for each presidential slate in each State” and “add such votes together to produce a ‘national popular vote total’ for each candidate.” That official, in each NPV state, would proclaim a national winner and appoint the slate of presidential electors pledged to that presidential candidate.

That’s a lot of power vested in a single official. There are some guardrails, of course—the compact directs the official to obtain other states’ vote totals from “an official statement containing the number of popular votes in a state for each presidential slate” if such a document is available (there are circumstances where it might not be—see “Electoral Count Reform Act poses serious problems for National Popular Vote” and “Missed it by that much (Part I)”).

Even when using an “official statement”—typically a state’s Certificate of Ascertainment, though other options include a statewide canvas or similar document produced at the end of a state’s election process—there are serious problems with trying to aggregate vote totals across state lines. Each state has its own way of casting, counting, and reporting votes that don’t always mesh well with NPV.

New York, for example, has not included tens and even hundreds of thousands of votes on its last four Certificates of Ascertainment, and Alaska’s and Maine’s ranked choice voting systems are particularly problematic for NPV because it produces more than a single vote total for each candidate.

Then there’s the problem of California’s 2016 Certificate of Ascertainment. A questionable ballot design decision made by Golden State election officials brings back memories of the Palm Beach County “butterfly ballot” in the 2000 presidential election, and could have flipped the expected 2016 outcome had NPV then been in effect.

Most people assume that Hillary Clinton won California in 2016, since she received 8,753,788 votes compared to Donald Trump’s 4,483,810 votes. Those are, in fact, the numbers on California’s 2016 Certificate of Ascertainment for the Democratic and Republican slates of electors, respectively.

But there was another slate of electors also pledged to Donald Trump in California that year, those of the American Independent Party (AIP). History buffs might recall that the AIP was the vehicle for George Wallace’s segregationist presidential campaign in 1968. Since its founding in 1967, it has always run its own candidate.

Except for 2016, when it decided to support Donald Trump.

According to a CNN story, the AIP approached the California Republican Party about running a common slate of electors (as happens in states with fusion voting, like New York) but the request was rejected. So the AIP ran its own slate of electors pledged to the Trump-Pence ticket.

Here’s where that questionable ballot design decision was made. Presumably in order to avoid confusing voters unaccustomed to seeing the same candidate listed twice for different parties (California does not have fusion voting), California ballots listed both the Republican and the American Independent designations next to the Trump-Pence ballot line (just as Hillary Clinton and Tim Kaine had the Democratic Party designation next to their names). This meant that every Californian who voted for the Trump-Pence ticket was voting for both the Republican and American Independent presidential slates, having in effect been given two votes for president.

This particular design decision didn’t have any impact on which candidate won California in 2016, of course. The Republican and American Independent elector slates technically tied for second place in the state, each with 4,483,810 votes. Because like most states California is “winner take all,” the Clinton-Kaine ticket received all 55 of the state’s electoral votes.

But the language of the compact is clear – the chief election official in each member state is to “determine the number of votes for each presidential slate in each State” and “add such votes together to produce a “national popular vote total” for each candidate.” The 2016 Certificate of Ascertainment for California shows 4,483,810 cast for the Republican Party’s slate of electors, and 4,483,810 additional votes cast for the American Independent Party’s slate of electors, both pledged to the Trump-Pence ticket. That totals to 8,967,620 votes cast for the Trump-Pence “presidential slate” in California in 2016.

Had NPV been in effect in 2016, California’s odd ballot design would have doubled Trump’s vote total in that state, giving him 4,483,810 more votes. That would have been more than enough to surpass Clinton’s 2,868,691 national vote margin. Under the naïve national vote tabulation scheme imposed by NPV, Trump won by 1,615,119 votes.

Could this actually happen if NPV was in effect? Wouldn’t election officials use common-sense and good-faith discretion to prevent such an absurdity when tabulating the votes? There are two reasons not to be so optimistic.

First, representatives of the organization that wrote and advocates for NPV claim that chief election officials in compact member states have no discretion at all. According to them, the vote totals on a Certificate of Ascertainment (or other “official statement”) are “conclusive,” leaving no room for discretion.

Second, even if NPV advocates are wrong and officials can exercise their discretion, it is not clear that every chief election official will always have sufficient common-sense or a desire to act in good faith. Members of both major parties have recently accused officials on the other side of manipulating the rules for partisan gain. For example, many Democrats have expressed grave concerns over “election denier” candidates, including those running for Secretary of State in a number of states, whom they fear might manipulate the 2024 presidential election process. Under NPV, opportunities for manipulation would increase. And, as shown above, in some cases no manipulation would be needed, only a willingness to stick by the precise terms of the compact in order to produce a clearly outrageous outcome.

A strict reading of the National Popular Vote compact, coupled with the unusual process used in California, would have declared the Trump-Pence ticket the popular vote winner in 2016. This is yet another example of how the National Popular Vote compact is fundamentally broken and should be abandoned.

Photo by Gage Skidmore