In two of my last three Verdict columns (this one and then this one co-authored with Jason Mazzone), I have discussed—in connection with the Supreme Court’s consideration in Trump v. Anderson of Colorado’s power vel non to exclude Donald Trump from the presidential ballot on account of his having been found by Colorado courts to have engaged in insurrection within the meaning of Section 3 of the Fourteenth Amendment—how Presidential elections are inherently decentralized. This decentralization inevitably leads to different election rules in different states, such that what a single state does can substantially affect who wins the White House whether other states (and their voters) like it or not.
The past week has brought additional vivid examples of this potent decentralization. In Ohio, a question has arisen whether President Joe Biden (or any ultimate presidential nominee of the Democratic Party) will qualify for access to the general-election ballot since the Democratic Convention won’t formally pick a party nominee until after the deadline has passed that Ohio currently imposes for major parties to name their nominees to ensure ballot access. An Ohio law mandates that party nominees, in order for their names to be included on the general election ballots, be certified at least 90 days before the general election. This year, that deadline to certify falls on August 7, but the Democratic National Convention, at which the party’s nominee will formally be selected, does not even begin (much less conclude) until 12 days later.
Even though Ohio is likely, after the November election is over and done, to give all of its presidential electors to the Republican candidate in any event (because Ohio is no longer a swing state but an increasingly reliable Red state), for Biden not to be on the ballot there would be a symbolic blow, reduce his national-popular-vote total (something people look to as a marker of legitimacy) and also probably hurt down-ticket Democratic candidates. In spite of these consequences, there is (rightly) general agreement that Ohio is entitled to have such a law as this (even if Ohio lawmakers know that such a deadline might ensnare one party or one candidate in particular in a given election) provided the law is formally equally applicable to both major political parties. And this would be true even if Ohio were a key swing state (as it has been in many past elections) whose outcome could tip the electoral college balance one way or the other.
The second recent episode involves Nebraska. Lawmakers there are considering a proposal (backed by Republican candidate Donald Trump) to alter the way the state allocates electors based on the November election results in the state. Right now, Nebraska (along with Maine) does not allocate presidential electors in a winner-take-all fashion. Under winner-take-all regimes, if a presidential candidate earns more votes than any other candidate in the state, that winning candidate earns the pledged support of all that state’s electors; a winner-take-all jurisdiction does not split its electoral votes between the various presidential tickets.
The prevalence of this winner-take-all approach to the electoral college around the nation should not come as any surprise if we take as a premise each state’s desire to maximize its own importance in the presidential election process. By providing each presidential candidate with a large return (in the form of the state’s entire electoral college bloc) for the candidate’s promises and platform planks targeted to the state’s electorate, the state increases the likelihood that all candidates will take the state seriously and address its needs and concerns.
But Nebraska currently (and historically) has eschewed a winner-take-all approach in favor of a district-by-district approach. Nebraska has five presidential electors (because it has two senators and three congressional districts). Under the current rules, three of the Nebraska’s electors would be awarded according to which presidential candidate wins the most votes in each of the state’s three congressional districts, with the remaining two electoral votes going to the candidate who wins the most votes statewide.
Such a district-by-district approach can allow a minority party (in Nebraska, the Dems) to nonetheless earn one or more of the state’s electors in spite of the state’s overwhelming redness statewide (because, say, Dems are concentrated in one urban congressional district). This happened in 2008 and again in 2020; the Republican candidate (John McCain and Donald Trump, respectively) won the statewide vote handily, but the Democratic nominee (Barack Obama and Joe Biden, respectively) picked up one of the five electors, for having won one of the three congressional districts.
This is not the first time Republicans have tried to revise Nebraska’s method of elector allocation to enhance the prospects for the Republican party. For example, similar efforts were made prior to the 2012 election. At that time there was also an effort to get Pennsylvania, a winner-take-all state that was likely (at that time) to vote for the Democrat in the presidential election but one that also had an elected state legislature controlled by Republicans, to move to a district-by-district method; had Pennsylvania gone from a winner-take-all to a district-by-district approach, Republicans might have picked up a significant number of electors (on account of winning several congressional districts). Indeed, especially because of partisan gerrymandering, it was possible for Republicans in the state to garner more than half the state’s electors (if they won enough congressional districts) even if the Democratic candidate got more votes (that were packed into a smaller number of congressional districts) statewide.
Whatever one thinks of the virtues of winner-take-all vs. district-by-district as a general matter, it would be good if states decided on their methods of presidential-elector allocation without regard to partisan outcomes (and it would be similarly good if states didn’t change their rules each election cycle based on partisan predictions). But whether proposals to change the method of elector allocation are motivated purely by partisan zeal in ways that strike many as unseemly or unfair (as did the proposal in Pennsylvania in 2011), they are not open to federal constitutional challenge, because Article II of the Constitution allows each state to appoint presidential electors more or less any way it chooses. Thus, like the Ohio-specific deadline for ballot access, state-specific (and thus disuniform) rules concerning elector allocation will persist under our decentralized electoral college framework.
A third reminder (for me) this week of the extent of decentralization was the mention by a colleague and frequent co-author (Professor Evan Caminker) of the outcome of the 2020 Texas v. Pennsylvania case in which Texas and other states filed for review in the U.S. Supreme Court invoking the now-discredited “Independent State Legislature” (ISL) theory to challenge the decision of Pennsylvania courts to enforce during the 2020 presidential election the state constitution over the state election code. Although the U.S. Supreme Court didn’t address and debunk ISL on the merits (as it later did last summer in Moore v Harper), the Court dismissed Texas’s filing on the ground that Texas lacked standing under Article III because “Texas has not demonstrated a judicially cognizable interest in the manner in which another state conducts its elections.”
Each of these three episodes highlights how underexplained the Court’s Trump v Anderson decision last month was. As Professor Mazzone and I explained in our column, the Court’s decision came down to the Court’s seeming belief in the need for a meaningful level of ballot uniformity across the nation, given that people in all states have input in choosing the President, as well as the President’s role as chief executive for the entire nation. According to the Court, chaos would result if presidential candidates were deemed ineligible by some states but not by others such that voters in different states would face different choices on election day, and the winner might be a candidate who wasn’t even on the ballot in some states. Adding to the problem, the Court reasoned, different states would likely employ different procedural mechanisms and different standards to determine ineligibility. “The result,” the Court worried, “could well be that a single candidate would be declared ineligible in some States, but not others, based on the same conduct (and perhaps even the same factual record).” Further, the Court thought as eligibility determinations unrolled across an election season, there would emerge “[a]n evolving electoral map” that could “dramatically change the behavior of voters, parties, and States across the country, in different ways and at different times.”
And yet whether we are talking about ballot access (as both Trump v. Anderson and the Ohio deadline law involve), or how votes are counted and used to allocate electors (at issue in Nebraska), variations among states guarantee that voters in some states are not treated the same way as are voters in others states, even though who is elected President affects people in each and every state. Moreover, unfolding decisions by Ohio and Nebraska in the crafting and enforcement of their rules could easily “change the behavior of voters, parties and States across the country” in an evolving way.
To be sure, each of the episodes discussed above could arguably be constitutionally distinguished from the Section 3 question at issue in Trump v. Anderson. In Ohio, for example, the ballot access question doesn’t involve individual candidate entitlement but rather party entitlement. But query why that should matter—parties and their nominees are closely intertwined, and excluding a party has the effect of excluding its (and its members’) preferred candidate.
Ohio’s determination of when a party declares its nominee might also seem less discretionary than Colorado’s factual determinations concerning Mr. Trump, but many key ballot-access determinations in many states (concerning, say, signature validity, volume, and timing) are far from mechanical, and yet we still allow states to do what they want in this regard.
With respect to the Texas v. Pennsylvania dismissal, perhaps it is not technically inconsistent to say a state lacks a cognizable interest for purposes of Article III in how other states administer presidential elections, and also say a state’s voters are protected against the specter of chaos arising from interstate disuniformity, but there is clearly a tension there (especially in light of the court’s recent willingness to let states represent their voters and citizens (as in, for example, Massachusetts v. EPA.)
Perhaps the bigger distinction between the Ohio episode (along with the Texas v. Pennsylvania episode and the Nebraska situation as well), on the one hand, and the Trump v. Anderson case, on the other, is that the former all involve only a state’s exercise of power under Article II, over which the federal government has no supervisory authority provided for in the text of the Constitution, whereas the latter involves the Fourteenth Amendment, Section 5 of which confers federal implementation power. But as Professor Mazzone and I pointed out last column, this grant of federal power alone does not explain why states are cut out of the enforcement loop. Section 5 power of the federal government does not, for example, foreclose a state from providing remedies against state officials who violate the Equal Protection or Due Process Clauses of Section 1 of the Fourteenth Amendment.
The big question remains: if disuniformity in presidential ballot access (or presidential election administration more generally) is a big constitutional problem, then why does Article II continue to permit such consequential disuniformity as reflected in the examples above? Relatedly, why does Article I give states the power to regulate congressional elections in disuniform ways (unless and until Congress says otherwise), even though, as we have seen above, partisan regulation of congressional districts can, depending on presidential-elector-allocation methods, have profound impacts on presidential selection?
These are the questions the Court needed to answer, and yet never even posed. That is why even if (and it’s a big “if”) Trump v. Anderson’s outcome is defensible, it surely was not actually defended satisfactorily by the Justices.