As my fellow Verdict columnist Professor Michael Dorf explained in his most recent column, the U.S. Court of Appeals for the Tenth Circuit recently issued a decision (in Baca v. Colorado Department of State) holding unconstitutional Colorado’s attempt to remove and replace a “faithless” presidential elector from the state’s electoral college contingent. The elector (Baca) had voted for a presidential candidate (John Kasich) other than the candidate for whom the elector had pledged to vote (Hillary Clinton) before being picked as an elector—hence the term “faithless.” The Tenth Circuit held that the “faithless” elector’s vote for Kasich was valid, notwithstanding the faithlessness.
The Tenth Circuit’s bottom line is plausible; the word “elector,” in the parts of the Constitution describing the electoral college, gives formal power to the individual casting a ballot, and not necessarily to the state that chose him, to determine whom to cast that vote for. In this brief commentary, I offer three observations that put the Tenth Circuit’s ruling in context.
First, even if the Tenth Circuit is correct (and would be affirmed by the Supreme Court) on the question whether an electoral college vote once cast must be counted, the same emphasis on constitutional text employed by the Tenth Circuit—that the word “elector” gives power to the person casting the ballot—permits the states wide latitude in picking the electors themselves. Article II gives power to “each state” to appoint (i.e., select) electors in “[any] manner” prescribed by the “legislature thereof.” The Supreme Court has repeatedly, and as recently as 2015, (rightly) made clear that the word “legislature” (outside the context of binary choices such as the ratification of proposed federal amendments) means any entity empowered under the state’s constitution and laws allocating lawmaking power. That means that Article II’s reference to “legislature” concerning the selection of presidential electors, like Article I’s reference to the power of state legislatures to determine the times, places, and manner of congressional elections, permits states to use the initiative and referendum devices that are part of state law, notwithstanding that states are performing a federal task in all these areas. As Justice Ginsburg wrote for a majority four years ago in Arizona State Legislature v. Arizona Independent Redistricting Commission, “Nothing in [Article I, whose language mirrors Article II’s] instructs, nor has this Court ever held, that a state legislature may prescribe regulations on the time, place, and manner of holding federal elections in defiance of provisions of the State’s constitution.” That case was explicitly relied on by a conservative majority in this past year’s Rucho v. Common Cause partisan gerrymandering case. The Tenth Circuit opinion in Baca recognized this, quoting Supreme Court authority for the proposition that the power of states to appoint presidential electors is plenary.
What does all this mean for the question of faithless electors? It means that states, if they want, can reduce elector faithlessness when electors vote by more carefully vetting the electors who are appointed in the first place. Right now, most states allow each political party to select a slate of electors who agree to pledge support for that party’s presidential and vice presidential candidates far in advance of the general election. This time lag increases the prospect of elector faithlessness months later. Instead, a state could allow the political campaign of each major party’s general election candidate to choose and offer a slate of electors (who are presumably highly loyal to that particular candidate) on the eve of the general election, and then select, based on the election results, the slate of electors associated with the winning campaign. Party loyalty is one thing; campaign and personal loyalty are far stronger, and the presidential candidates themselves are very unlikely to include would-be electors who pose any real risk of faithlessness.
Second, even after electors are appointed, and prior to their formal votes for President and Vice President, states might be able to better police faithlessness. The essence of the Tenth Circuit’s holding is that a vote once case cannot be undone. But what if a state had a law that said something like the following:
The electors of the President and Vice President selected by the voters of the state on the date prescribed by federal law as national election day shall convene at the seat of government on the day fixed by federal statute, at the hour of twelve o’clock noon of that day. The electors shall perform the duties required of them by the Constitution and laws of the United States by voting separately for President and Vice President. Shortly before their vote, they shall take a straw poll, with each elector announcing her or his preferred presidential and vice presidential candidates. After a straw poll of the electors is conducted, it shall be deemed to be the final vote of the electors if no elector objects within 30 minutes. But if any elector objects to deeming the straw poll results to be a final vote, a vote shall be taken in which a majority shall decide whether the previous straw poll should be considered a final vote. Once a vote is deemed final, its results shall be certified to the relevant state authorities. But prior to a final vote being accomplished, the electors, like each house of Congress, may expel members of the electoral group for bad behavior, including refusal to act or refusal to act faithfully, and may fill any vacancy in the office of an elector occasioned by death, refusal to act, refusal to act faithfully, neglect to attend, or otherwise, by voice vote, and plurality of votes.
Perhaps the reasoning of the Tenth Circuit opinion would call this into question too, but perhaps not, since in the Colorado case a vote by the electors had clearly occurred before the state tried to nullify it. There might be some tension between my proposal and the timelines set up in a federal statute passed in the nineteenth century to deal with disputes over the appointment of electors, but my proposed statute does not, strictly speaking, relate to disputes over appointment so much as removal of persons who were appointed. Moreover, the federal statute does not tell states what to do; it merely provides that any dispute resolutions that states reach in particular ways at particular times will be respected by Congress when it tallies electoral college votes. It has never been clear whether congressional vote talliers would be obligated to follow the federal statute anyway. And if my proposed solution satisfies Article II of the Constitution, states needn’t worry about a statutory “safe harbor.” Finally, if there is tension between my proposed solution (or other proposed solutions) to the faithless elector problem and federal law, federal statutes should be rewritten. Indeed, Congress should specify a time period—e.g., from national election day until two weeks later—during which electors are technically appointed, rather than specifying a particular single date (as current law purports to) for the actual appointment. Such statutory revision, unlike constitutional amendment, is within the realm of the politically feasible.
Third, and finally, if and when the National Popular Vote (NPV) Interstate Compact plan were ever to be approved by Congress and get up and running (more on that plan here and many prior columns), it would pave the way to a constitutional amendment fully replacing the electoral college with a straightforward national popular election for President. I do not share Professor Dorf’s concern that the Tenth Circuit’s opinion would undermine people’s confidence in the NPV plan’s validity (and some people think the Tenth Circuit’s decision helps NPV), for three reasons. First, as noted earlier, the text of the Constitution, which might cut against Colorado’s efforts to date to regulate faithless electors, supports, not undermines, the power of states to appoint electors any way they choose (including by means of NPV). Second, Congressional approval and enforcement of the plan (which I think is essential given the way the plan has been crafted) would blunt any fear of states not following through on their NPV obligations. And third, NPV, once it crests over the 270 threshold, could be adjudicated in the Supreme Court prior to its use in an election, which again would provide clarity and reliability.
This two-step process—implementation of NPV followed ultimately by constitutional amendment—is patterned after other important changes in constitutional election procedures. Indeed, although many commentators and courts are completely (and shamefully) oblivious in this regard, this pattern of state-level innovation followed by federal constitutional amendment is precisely how direct election of U.S. Senators came to be entrenched in the Constitution. Analysts (including Chief Justice Roberts in his dissent in the 2015 Arizona case, a dissent from which he rightly backed off in Rucho) have said things like: “If states on their own could have changed the way federal elections are conducted, then we wouldn’t have needed the 17th amendment to accomplish direct election of Senators.” It turns out, although very few people are aware of it, most states, in the years prior to the submission of the 17th Amendment for ratification, had already—through devices very similar to the NPV concept—moved to direct election by requiring as a matter of state law that the elected legislature of the state honor the wishes of the state electorate in selecting U.S. Senators. Indeed, the U.S. Senate agreed to submit the 17th to the states for ratification only after many, if not most, U.S. Senators had already been directly elected themselves as a result of state-law innovation, and thus did not fear the direct election device. Justice Holmes once said a page of history is worth a volume of logic.” When history and logic coincide, the combination is more powerful still.