Who elects the President?

Under our Constitution, the President is elected by “electors.” There are currently 538 electors. To win, a candidate needs a majority of all electors appointed, so 270.

State legislatures decide how electors are chosen. Initially, many legislatures chose the electors directly themselves. Since 1876, every legislature had given the voters in their state the presumptive choice of electors through an election. 

When people vote for the President on this year’s November 3 election day, the results of the popular vote in each state determines which party sends its slate of electors to represent that state in the Electoral College. How electors are chosen differs from one state to another. On December 14, those “electors” then gather in their state to cast their vote for President. 

The electors’ votes are then transmitted to Congress. On January 6th, the new Congress, which is seated on January 3rd, is to gather in a Joint Session to count those votes. The candidate who receives a majority (270) is elected President.

How then does the popular election on November 3 election work?

In the ordinary case, the people vote on election day and, often, before election day, either by mail or through “early voting” that happens before November 3. But November 3 is the cut-off: it’s the last day to vote, and, when the polls close in a state, the results may start to be released. 

Each state has a procedure of regulating and policing the count. (You can see a table listing those procedures here.) After the count is completed, the winner is certified and the electors associated with the winning candidates are appointed to vote in the electoral college. In all but two states (Maine and Nebraska), the winner of the vote in a state gets all of the electoral votes from that state. In Maine and Nebraska, the winner of each Congressional district gets one elector, and the statewide winner gets two.

What is the “Red Mirage”?

The “Red Mirage” (also known as the “Blue Shift”) is a potential 2020 election scenario in which President Trump will appear to be ahead on Election Night when the returns are first reported, only to then fall behind  — and ultimately lose to — Vice President Biden as all the mail ballots are counted in the days (and potentially weeks) after November 3rd.

This outcome is particularly concerning because the progression of the overall count would superficially appear to coincide with the President’s rhetoric about the Democrats’ plan to use mail-in voting to “steal the election”.

In reality, this “red mirage” / “blue shift” need not be the result of any sort of maleficence or fraud. It is instead explained by three key factors:

1) historic levels of mail voting,

2) election laws in several crucial swing states (PA, MI, WI) that do not provide sufficient time for the counting/processing of mail ballots before election day, and

3) a distinct partisan split between “by mail” (overwhelmingly Democratic) vs. election day (overwhelmingly Republican) voters.

1. Mail Voting

In lieu of going to their polling place on election day, many registered voters in the United States also have the option to “vote early” either in person or by mail. This latter option is most commonly known as voting “absentee” or voting “by mail.” To be clear, every state sets its own rules related to election administration and as a result, there are a diversity of early voting policies used in different states throughout the country. For example, some states like Utah and Oregon use an entirely “vote by mail” program in which all registered voters are automatically sent a mail ballot, which is then filled out by the voter and returned by mail or physically dropped off at a specified location (election office, dropbox, etc.). Others only offer an absentee by mail option for voters who can provide a state-approved “excuse” for why they will be unable to vote in-person on election day. Finally, a number of states have substantially tweaked the original absentee concept, and allow for “no excuse” absentee voting, in which any registered voter can request a mail ballot without providing a reason. 

Following the outbreak of COVID-19, many states modified their absentee laws to permit more flexible standards for voting by mail. For the 2020 presidential election, only five states (Texas, Louisiana, Mississippi, Tennessee, Indiana) will still require an excuse to vote absentee. Thus, given the COVID related health dangers posed by in person voting for many citizens this fall, and the related widespread loosening of certain absentee voting requirements, the country is posed to vote by mail in truly unprecedented numbers in the upcoming 2020 election. In 2016, 33 million Americans voted by mail; in 2020, that number is projected to skyrocket to over 80 million.  Notably, the unprecedented increase in the quantity of mail-in ballots –and some states’ failure to adjust procedures accordingly — will serve as one of the main drivers in any Red Mirage scenario.

2. Counting/Processing Mail Ballot Laws in WI, MI, PA

In 2016, President Trump defeated Hillary Clinton in Michigan by 0.2%, Pennsylvania by 0.7% and Wisconsin by 0.8% — winning these three states by 10,704, 46,765 and 22,177 votes respectively.

In 2020, if the Presidency again hinges on close races in MI, WI, and PA:

  1. We won’t have a projected winner on election night
  2. If Biden wins, then there will likely be some sort of red mirage that is significantly featured during election “week”

Why? As things currently stand, Wisconsin, Pennsylvania, and Michigan are all slow counting states — most notably due to the fact that they either only allow less than 1 day (MI), or do not allow any (WI/PA) pre-election day processing of mail ballots.

Previously, the lack of early processing time in these states has not posed a significant problem — in 2018, around one quarter of the votes in Michigan — and only around 5% of all votes in Wisconsin and Pennsylvania — were cast via mail ballot. However, due to the impact of COVID-19 discussed above, mail ballots are projected to make up around 50% of the total votes received in WI, MI, and PA this fall. Although public officials have long known that absentee voting demand would skyrocket for the upcoming Presidential election, state legislatures in these three states have each failed to sufficiently update their counting and processing laws to properly account for this reality.

In Michigan, lawmakers recently moved back the processing start date from November 3rd at 7am to November 2nd at 10am in cities or townships with at least 25,000 residents. Although this is an improvement over simply preserving the status quo, it falls well short of the seven day minimum head start that the Bipartisan Policy Center’s Task Force on Elections recommends, or the two weeks+ many of their fellow swing states (ex: AZ, NC, FL, etc.) permit. Consequently, Michigan will not have a declared winner on election night — state officials have expressed confidence that they will have “unofficial statewide results” available by November 6th.

In Pennsylvania and Wisconsin, election administrators still may not begin processing mail ballots until election day . So it could be at least a few days after November 3rd until all the ballots get counted in these two pivotal states. Interestingly, officials in WI have expressed optimism that they will have results on November 4th — Gov. Evers recently remarked that he “believe[d] that we will be able to know the results of the Wisconsin election, hopefully that night and maybe at the latest the very next day”. In Pennsylvania, widely considered to be the single most important state in tipping the balance of the 2020 election, there is less optimism. In addition to their late start on on the front end of the counting process, PA absentee ballots can also be delivered as late as November 6th, further elongating the counting process. Thus, if things go like they did in PA’s primary earlier this summer, it may take multiple days, and potentially even longer, to get all the results in.

A Partisan Divide — Early Voting vs. Election Day

In addition to the exponential increase in the number of people voting by mail in November and the inadequate amount of time election administrators have to process/count these ballots in key swing states (WI/PA/MI), the distinct partisan split between those planning to vote by mail (Democrats) vs. on election day/in-person (Republicans) is the final essential factor in explaining the driving forces behind a potential “red mirage.”

According to an August survey of likely voters by NBC News:

Supporters of Democratic candidate Joe Biden are significantly more likely than Trump backers to say they plan to vote by mail. Nearly half (47%) say they plan to mail in their ballot, with an additional 21 % saying they will cast a vote before Election Day at an early in-person voting site. Only about a quarter of Biden voters, 26 %, plan to vote on Election Day at a polling place. In contrast, two-thirds of Trump’s voters — 66 % — say they will vote in person on Election Day. Just 11 percent say they plan to vote by mail, and 20 percent say they will vote early in person.

Source: https://www.nbcnews.com/politics/2020-election/siloed-nbc-wsj-poll-shows-how-gop-dems-hold-widely-n1236958

In September, a NYT/Siena poll identified a similar sentiment among likely voters in key swing states across the country: Just 26% of Democrats said they planned to vote in person on Election Day, compared with 56 % of Republicans, according to polling of likely voters in 11 battleground states.”

In a survey from September 30 – October 5th, Pew Research Center identified Trump voters as still more inclined to vote on election day, and Biden voters with a continued preference to vote by mail:

Trump supporters are more than twice as likely as Biden supporters to say they plan to vote in person on Election Day (50% vs. 20%). By contrast, 51% of Biden supporters say they plan to vote by mail or absentee (or have already voted this way). A quarter of Trump supporters (25%) say they plan to vote by mail or absentee.

Finally, a NPR/PBS/Marist poll conducted early-mid October :

“showed voters planning to vote in person supported Trump 62-35%, while those planning to vote by mail backed Biden 73-23%. And voters planning to cast their ballot at an early voting location favored Biden over Trump 65-32%. In all, the poll found, 65% of Trump voters intended to cast their ballots on election day in person. Biden voters were more split, with 43% planning to vote by mail, 26% at an early voting location and 29% in person on election day.

Source: http://maristpoll.marist.edu/wp-content/uploads/2020/10/NPR_PBS-NewsHour_Marist-Poll_Likely-Voters_USA-NOS-and-Tables_202010141328-1.pdf#page=3

This partisan outlook on voting by mail vs. voting on election day has not only emerged as a consistent theme in public polling, but has also been supported by actual early voting returns.

To be clear, both parties have actually pushed their respective bases to consider alternatives to their general preferred methods of voting in order to either avoid “lost” votes (Ds) or boost turnout (Rs). For those on the left, concerns related to USPS delays and increased rates of rejected absentee ballots has led Democratic party leadership to encourage voting early in person if possible. On the other side, Republicans have sought to reassure and encourage skeptical GOP voters that it is in fact safe to cast an absentee ballot — a tricky task given the President’s barrage of unsubstantiated attacks on the practice.

Yet, the overall partisan dynamics still hold — put simply, Biden voters are overwhelmingly more likely to vote prior to election day, while the majority of Trump voters prefer to do so in person on November 3rd. As a result, should the election come down to “slow-counting” states like Pennsylvania, Michigan, and Wisconsin, the initial returns would likely appear to show Trump in an advantageous position to win the election based on election day turnout. However, as the mail ballots are counted in the days that follow, the President’s lead would evaporate and potentially lead to a Democratic victory. 

Drawing all this together, a Red Mirage will be caused by:

1) a historically high quantity of mail-in votes;

2) states’ inability, or intentional reluctance, to amend their laws related to allow for sufficient pre-election day processing/counting of mail-in votes; and

3) the electorate’s polarized preferences for mail-in vs. in-person voting.

A brief summary of this concept, applied to the upcoming election, is provided below.

If Biden needs WI, MI, or PA to get to 270, then a red-mirage is likely.

The mail-in votes in these states are skewed heavily democratic, and the in-person vote skew Republican. Due to the slow processing/counting dynamic discussed above, the counts displayed on election night will likely show Trump initially leading by a decent margin. In order to further undermine the legitimacy of the voting process, the President and partisan news stations may also try to initially claim victory based on this initial lead on election night — despite knowing that many votes remain uncounted. However, as all the mail in votes are counted, Vice President Biden’s numbers will continue to improve, and will probably be large enough to overcome the deficit presented in the initial returns. 

Can a legislature change its mind and cancel an election while it’s ongoing or after it happens?

It’s never happened and we think it would be illegal.

Well before an election begins, a legislature could decide to remove the people from the process that selects electors. The last state to flip flop like this was Florida in the 19th Century. In 1860, it held a popular vote for President. But after Florida seceded during the Civil War it didn’t participate in the presidential election of 1864. After Florida rejoined, it passed a law well before the 1868 election allowing its legislature to directly appoint electors. Florida then reversed course again before the 1872 election and held a popular vote for President. It has held a popular vote for President ever since. But, in theory, Florida could pass a law before the 2024 election and decide its legislature should appoint electors directly.

But after an election has begun or after it’s been held, however, it’s less clear that the state may retroactively change the method of appointing electors. The best authority suggests it could not. 

First, such a procedure would seem to violate 3 U.S.C. § 1, which requires that states that hold elections select their electors on election day. If a legislature attempted to change the slate of electors after Election Day, it would miss this deadline and violate federal law.

Second, there’s a strong argument a state legislature’s substitution of its own slate of electors after Election Day would violate the Due Process Clause of the Constitution.  The Due Process Clause protects citizens’ expectations under state law that they will be able to exercise their right to vote and have that vote counted.  If the legislature elevates its own preferences over those of the voters by retroactively changing state law, this right would be violated.

Both limitations would depend, however, upon the courts intervening. It is possible the Courts would view the legislature’s actions as a “political question,” not subject to judicial review.

However, a state could claim that the election has “failed” under the “Failed Election” provision of Title 3.  As discussed, below, states could cite this provision as an authorization for sending their own slates of electors to Congress.

What if the election has “failed” under the “Failed Election” provision of Title 3?

Congress has gathered all the statutory provisions related to the election of the President into Title 3 of the United States Code. Some of these provisions were part of the Electoral Count Act of 1887. The “failed election” provision of Title 3 was enacted 30 years before the Electoral Count Act, at the time when Congress set election day to be a single day rather than a range of days during which the votes might occur.

3 USC §1 declares the day on which the presidential election shall happen. §2 declares:

Whenever any State has held an election for the purpose of choosing electors, and has failed to make a choice on the day prescribed by law, the electors may be appointed on a subsequent day in such a manner as the legislature of such State may direct.

The language “failed to make a choice” has created some confusion. We know the language was crafted to deal with a very specific problem: certain states (eg. New Hampshire) had a requirement that the choice of an elector for President had to be by a majority, not a mere plurality. This provision gave the states the ability to keep such a system, but then provide for a backup election when such a requirement was not met. 

Today, there are a number of states that have implicitly taken advantage of this provision in the case of a tie.  The is one state, North Carolina, that expressly references this section in the case when no electors are selected by the time in which they must vote.

Whether 3 USC § 2 could apply to other causes of failure is a harder question.  If the “failure” to make a choice were applied to something other than a majority requirement— such as an emergency caused by a natural disaster—a state might argue that Section 2 gives it the power to pick its own slate of electors in the face of this other types of “failure.” Applying the “plain meaning” of the term “failure,” this interpretation seems possible, even if it reaches far beyond the original context that drafters of this provision had in mind.   

It seems that any of the states can send additional slates of electors to Congress.  Under the ECA, it is the President of the Senate who opens the slates of electors.  The crucially important point is that the slate must be signed by the state’s governor to be valid.  This is where political party might matter.  It is less likely that a state with a Republican legislature and a Democratic governor could pull off an alternative slate; the governor is unlikely to sign the slate.  However, in a state with both a Republican legislature and governor, it seems more likely that an alternative slate would be certified.  

Section 15 of the ECA governs what happens next.  3 U.S.C. § 15.  If more than one slate gets sent to Congress and neither is presumptively valid against the other, the Joint Session must pick which to count.  If both Houses agree, that slate is counted.  But if they don’t (Congress is tied), the slate that is certified (signed) by the governor is the one that prevails.  

The Original Intent of 3 U.S.C. § 2: Give Judiciable Power to State Supreme Courts in Cases of Election Disputes

3 U.S.C. § 2 specifies that if a state’s election fails to choose a slate of electors, then the state legislature can submit a slate of electors otherwise. The original thrust of the statute was to let each State Supreme Court be in control of determining each state’s determination in cases where the election result is indeterminable. See generally Edward B. Foley, The Analysis and Mitigation of Electoral Errors: Theory, Practice, Policy, 18 STAN. L. & POL. REV. 350 (2007). See also Michael Pitts, Heads or Tails?: A Modest Proposal for Deciding Close Elections, 39 CONN. L. REV. 739 (2006). Congress was especially concerned about the problem of a “statistical tie,” in which case the presidential result would be determined essentially by random. Giving the states the power to determine results gives State Supreme Courts the power to adjudicate. While on its face, the statute appears to give state legislatures the power to determine electors, a part of the original legislative intent was to give power to the State Supreme Courts. Principles of Election Law § 317 (2019).

Legislative History of 3 U.S.C. § 2: Applied Only During Unforeseen Circumstances

On its face, the statute gives wide latitude for the State legislatures and State Supreme Courts for deciding when to apply the statute. However, in the legislative history of the debates about 3 U.S.C. § 2, Congress seemed to have intended the statute’s application to apply only in circumstances of “unforeseen emergencies.” Under such emergency circumstances, then the effects of 3 U.S.C. § 2 would apply, and if no such emergency arises, then 3 U.S.C. § 2 would not be applicable. If such an emergency arises, then the discretion is left to the State Supreme Courts and state legislatures, as previously analyzed. 

The next question is, whether states’ emergency statutes would apply, or if the effects of 3 U.S.C. § 2 would only apply in an extreme, narrow set of cases. The courts have not addressed this particular issue. Some scholars, like Michael Morley, postulate that 3 U.S.C. § 2 can only be properly invoked for a narrow set of extraordinary circumstances: devastating natural disasters or terrorist attacks that incapacitate abilities to vote count. See Michael Morley, Election Emergencies: Voting in the Wake of Natural Disasters and Terrorist Attacks, 67 Emory Law Journal 545 (2018). These types of emergencies do not seem to align with the type of emergency envisioned for the 2020 election. In the past, states have invoked emergency statutes for issues ranging from fires burning all of the vote counting machines in a county to hurricanes incapacitating a state’s ability to exercise its election processes. None have contemplated the types of mail-in ballot fraud issues that President Trump and others are alleging, and it is possible that courts would find this type of situation to constitute an “extraordinary emergency,” even if this type of emergency does not conform to previously envisioned situations. Moreover, because 3 U.S.C. § 2 has not yet been invoked before—Bush v. Gore sidestepped the potential issues of a state changing out a slate of electors—it is foreseeable that states would want to take advantage of this sort of unprecedented emergency allegation to invoke 3 U.S.C. § 2 to choose their own slates of electors. In such a circumstance, there are two likely scenarios to arise.

Unfortunately, because 3 U.S.C. § 2 is worded to give states latitude in making decisions when they fail to make a choice during an election, some states have adopted their own statutes in ways that liberally interpret, and in some cases, even go against the original legislative intent of 3 U.S.C. § 2. North Carolina’s disputed election statute, for example, gives its state legislature broad authority to select slates of electors in disputed election scenarios, even if the grounds for such selections are shaky. Thus, North Carolina’s state legislature could theoretically call an emergency session to appoint its own slate of electors even if the evidence is not clear whether an extraordinary emergency has occurred and even if there is not good evidence to support claims of fraud or otherwise. These liberal derivative statutes, based on 3 U.S.C. § 2, ignore the legislative intent and present risks of rogue legislatures altering the election results by declaring their own slates of electors.

Two Scenarios

One likely scenario that might arise is that the election results may be postponed because of some states’ statutes that allow for postponement in case of emergency. Given that 3 U.S.C. § 2 gives states latitude for determining electors in cases of emergency, states following their own emergency statutes could postpone the election results for their state. 

The other likely scenario is the issue of a governor appointing an alternative slate according to the state’s emergency statutes. Many states have provisions in their state statutes that the governor would choose a slate of electors if the election is inconclusive or in dispute. Following these statutes, Republican governors in battleground states could potentially appoint a slate of electors to re-elect President Trump, even if the ballots tip towards Vice-President Biden.

Arguments Against Invoking 3 U.S.C. § 2 for Allowing States to Submit an Alternative Slate

Both of these scenarios are deeply unsettling—both could lead to chaos and major confrontation. Additionally, both raise fundamental questions about whether 3 U.S.C. § 2 should be invoked in such a situation. If either of such scenarios arise, fighting the applicability of 3 U.S.C. § 2 in federal and state courts becomes paramount. Specifically, the challenge should be centered around whether a ballot fraud allegation constitutes an emergency situation. 

The crux of the issue is that states have radically different ways of dealing with election disputes. Some, like Louisiana, allow delays in voting, others give the state executive full authority to remedy. In those states that give the governor the power to determine the slate of electors, determining whether the governor’s power is an appropriate remedy is the central question. Following precedent, state emergency statutes generally give the governor broad powers to declare a state of emergency and postpone ballot counting in order to take steps necessary to protect human life, but these have only been applied in cases where it was made physically impossible or difficult to cast a ballot, either by terrorist attack (like 9/11) or natural disaster (fire or hurricane). See id. These situations do not seem to be congruous with the potential challenges that this upcoming election poses. Mail-in ballot fraud or other types of election fraud do not constitute an immediate threat to human life. Arguing that the type of “emergency” raised in the 2020 election does not constitute the type of emergency envisioned by 3 U.S.C. § 2 or by past precedents is a potential way to argue against invoking the statute. 

Textually, the phrase “failed to make a choice” does not mean that the choice itself is difficult to ascertain. To “fail” means “to be unsuccessful” or “to fall short.” Meriam-Webster.com Dictionary. Voters themselves do not “fail to make a choice” unless an exactly equal number of voters vote for Biden as for Trump. In that case, voters would “fail to make a choice” because there was a tie. Under Florida law, for instance, a tie would be broken on a later date by the drawing of lots. Florida Statutes § 100.181. A tied election falls under the exception to election day appointment given in 3 U.S.C. § 2. A stalled election does not.

This is consistent with the Supreme Court’s interpretation of the parallel provision of federal law for Congressional elections, 2 U.S.C. § 8, which uses nearly identical phrasing. That provision provides that when there is a vacancy in federal office caused by a “failure to elect at the time prescribed by law,” then State law may prescribe an election on a subsequent date. 2 U.S.C. § 8(a). The Supreme Court has held that a “failure to elect” occurs where all votes are counted but, by law, no candidate has been elected—because, say, there is a requirement that a candidate receive a majority of the vote to be elected. Foster v. Love, 522 U.S. 67, 72 (1997).

In a footnote, the Supreme Court explained why 2 U.S.C. § 8 would permit a state to hold a run-off election if it held an election with a majority-winner requirement on election day and no candidate received a majority of the vote:

Title 2 U.S.C. § 8, which was enacted along with § 7, provides that a State may hold a congressional election on a day other than the uniform federal election day when such an election is necessitated “by a failure to elect at the time prescribed by law.” The only explanation of this provision offered in the legislative history is Senator Allen G. Thurman’s statement that “there can be no failure to elect except in those States in which a majority of all the votes is necessary to elect a member.” Cong. Globe, 42d Cong., 2d Sess., 677 (1872) (remarks of Sen. Thurman). In those States, if no candidate receives a majority vote on federal election day, there has been a failure to elect and a subsequent run-off election is required.

Id. at 72 n.3. As the Supreme Court recognized with respect to 2 U.S.C. § 8 in Foster, there is no doubt that 3 U.S.C. § 2 was enacted to permit states to conduct just a run-off election in the case of a tie.

Whether Courts Apply Legislative History is Key

If it would be difficult to make the case for an emergency in front of the courts, would 3 U.S.C. § 2 still be applicable? The legislative history would seem to indicate otherwise. Interpreting the statute on its face, on the other hand, there is a colorable legal argument that the statute would apply even absent an emergency—as long as there is a dispute, state legislatures can override and appoint their own slate of electors. The deciding factor comes down to the state courts—what Congress had originally anticipated when drafting the statute—to decide whether legislative history and legislative intent should factor into deciding whether 3 U.S.C. § 2 kicks in if a dispute arises in any particular state. 

Fraud and the Application of 3 U.S.C. § 2

Fraud cases and close re-election cases have been present before, but the case is different for presidential elections as opposed to local or congressional elections. One option available in congressional elections, for instance, is simply to postpone the election or seat the Senator/Congressman later, when another election or a recount is held. That option does not exist for the presidential election. Never before in US history has there been a second election held for the presidency, and the timeline for inauguration is firm—there is no time to re-hold the election.

Common People v. Georgia found a multi-prong test for testing whether emergency statutes would be triggered for elections. Among those prongs is common law fraud. If the Georgia case’s logic is to be followed, then there may be a case to be made for common law fraud to be the grounds upon which to sue in the event of an election law dispute. However, the elements for common law fraud—intentional misrepresentation, knowledge of the falsehood, reliance, and damage—do not seem to apply very well to election law. Moreover, existing election fraud statutes potential override any claims for common law fraud. An argument for terrorism/outside intervention induced failure to make a choice, however, seem to fall under the type of emergency that 3 U.S.C. § 2 anticipated. How courts would respond to such an argument is unclear.

Presidential election disputes are not rare, but past disputes came only within slim margins—a recount in 1960 for Nixon in Hawaii, Florida in 2000 for Bush—and none have even come close to the allegations that the Trump Administration is peddling. In fact, the Brennan Center at the New York University School of Law found only four cases of documented fraud in the 2016 elections. If, however, the Trump Administration challenges the election in courts on fraud grounds, there would simply be no time for a runoff election.

A Remedy: Do Not Count Dispute States in Final Tally

One remedy for seriously disputed elections is, instead of appointing an alternative slate of electors and go through the legal gymnastics of disputing that alternative slate, to simply not count the electoral votes of that particular state should the state fail to make a choice. If Florida fails to make a choice, then its electoral votes would not be accounted for during the final race to 270. The predicate of this type of remedy is that the court finds that there is no clear outcome to the election of that state. Constrained by time, this would perhaps be the most sensible resolution in case a serious dispute arises in any given state. 

What if there is an emergency on election day? Can the legislature override the election then?

Currently, there is no federal law addressing the process for dealing with emergencies on Election Day.  Some state laws provide guidance for Election Day emergencies, however.  Generally, these laws outline procedures for postponing or suspending elections.  The most common statutory approach is to permit delaying the election, changing polling place locations, or doing both.  A state-by-state chart is shown here.

California, Florida, Oklahoma, and Virginia have the most expansive emergency election statutes.  Most emergency statutes require unilateral executive action, meaning the governor is given the power to delay or postpone an election.  Several states require either direct gubernatorial involvement or consultation with the executive branch in an election emergency (Florida, Kentucky, Louisiana, Maryland, Oregon, South Carolina, Texas, Virginia, Utah, Illinois, and Maine).  Under these state law regimes, therefore, it seems the governor has significant power to delay or postpone elections based on either a direct statutory grant of power or indirectly by declaring a state of emergency.

New York has a unique statute that allows for another day of voting not more than 20 days after the original election day if an emergency causes fewer than 25 percent of registered voters to show up to vote.  NY CLS Elec. § 3-108.

Events that can trigger emergency statutory provisions include public health crises, floods, hurricanes, earthquakes, fires, power outages, active shooter situations, cyberattacks, or any number of other “emergencies.”  Most statutes define “emergency” broadly, and states likely have significant power to manufacture emergencies based on Covid-19 or anything that might reasonably be termed an emergency.  However, they would need unilateral support from the state executive branch, and under most state laws would need to provide an alternative mechanism for voting.

Whether a state law that authorizes a state official to change the day on which an election is held conflicts with federal law (3 USC § 1) is an open question. It is possible that such a law would be interested as an application of 3 USC § 2, and be deemed the alternative procedure for selecting electors because of that “failure.”

What happens if there is more than one “slate” of electors purporting to represent a state?

It has happened in our history that there has been more than one slate of electors that purports to represent the state. The most famous example is 1876, when four states sent two slates of electors to Congress to be counted. But it also happened in 1960, when Hawaii sent more than one slate to be counted because a recount of a very close vote changed the winner from Nixon to Kennedy. 

The 1876 incident inspired Congress to pass the Electoral Count Act of 1887 (“ECA”). That act aims to regulate the process of counting electoral votes. The ECA offers a procedure — complicated and incomplete — for determining which electoral votes will be counted. 

The most important part of this procedure is the “safe harbor” provision of §5. (Note, we’ll refer to the current section numbers within Title 3 of the U.S. code, even though they weren’t the original section numbers. As we’ve said, Title 3 combines a bunch of statutes passed at different times.)

Under §5, in order for a slate of electors to qualify for “safe harbor,” the electors must be chosen according to the laws of the state enacted prior to Election Day (November 3), and any disputes related to the appointment of electors must be resolved at least 6 days prior (December 8) to the meeting of electors (December 14). 

If between two competing slates of electors, one of them qualifies for “safe harbor” protection, then that slate of electors will be counted (unless both chambers of Congress reject the vote). 

However, if there are multiple slates from a state that qualify for “safe harbor” protection, and if both the chambers agree to accept a particular slate, then that slate would count. Similarly, if both chambers agree to reject a particular slate, then that slate would not be counted. 

If both chambers cannot agree to either accept or reject a slate, then the slate which was certified by the ‘executive of the state’ (generally considered the governor, but some interpretations would also consider Secretary of State to be an executive too – thus adding to the uncertainty) gets counted.

What if a slate doesn’t get “safe harbor” protection?

According to the Electoral Count Act, in order for a slate of electors to qualify for “safe harbor”, the electors must be chosen according to the laws of the state enacted prior to Election Day (November 3), and any/all disputes related to the appointment of electors must be resolved at least 6 days prior (December 8) to the meeting of electors (December 14). 3 U.S.C § 5. 

Congress usually shows deference to slates qualifying under the “safe harbor” protection, and as long as there is a single slate from a particular state, and there is no allegation of corruption, the slate would be counted.

A slate would fail to qualify for the “safe harbor” protection, if either the electors were appointed in contravention to the election laws enacted prior to the Election Day,  or if the disputes were not resolved six days prior to the meeting of the electors. 

If Congress does not object to counting the votes from a slate that did not qualify for “safe harbor” (provided there was a single state presented from the state), there is no problem and the votes get counted. However, if Congress objected to counting votes from a slate that did not qualify for “safe harbor,” and there are multiple slates from a state, then the Congress would have to decide on which slate to count.

What if more than one slate claims safe harbor protection under different state procedures?

If there are multiple slates from a state that qualify for “safe harbor” protection, and if both the chambers agree to accept a particular slate, then that slate would count. Similarly, if both chambers agree to reject a particular slate, then that slate would not be counted. 

However, if both chambers cannot agree to either accept or reject a slate, then the slate which was certified by the ‘executive of the state’ (generally considered the governor, but some interpretations would also consider Secretary of State to be an executive too – thus adding to the uncertainty) would be counted. 

This determination of accepting or rejecting a slate of votes must be carried out in support of the State law and must respect the decision of the state’s election dispute resolution body (usually a court). The chambers may reject a slate if the electoral votes were ‘properly given’ – that is, there was no forgery, corruption, votes were cast by qualified electors who were voting on the correct day, electors voted for the correct candidate and complied with state laws that bind the electors.

How does Congress decide these questions?

The count happens in a Joint Session of Congress, presided over by the President of the Senate (Vice President Mike Pence). 3 U.S.C. §15 outlines the procedure. When the count begins, Pence will start with Alabama and proceed in alphabetical order. 

As each state is counted, members of either House have a chance to object. Any objection must be filed in writing, without argument, and signed by at least one member of the House and one member of the Senate. 

If the objection is then received, the procedure says that the Senate has to return to its chamber, and each House decides whether to support the objection (sustain it) or oppose it (overrule it). Once both have decided, the Joint Session resumes. The vote of each House is then recorded. 

Can Vice President Pence, in his capacity as the President of the Senate, pick and choose which electoral votes to count? Probably not. 

Twice in American history, the President of the Senate has both been a candidate for President, and decisively used his powers during the counting of the vote. In 1797, John Adams counted a decisive slate of votes from Vermont which had been questioned in the spring. And in 1801, Thomas Jefferson similarly counted a decisive slate from Georgia which were legally defective on their face.

In neither case did the opposing party object to decisions undertaken by the presiding candidate. While there is clear historical precedent that an interested President of the Senate may exert influence over the count, his powers are limited, particularly where an opposing party is motivated to contest his decision.

What happens if there’s a tie between the Houses?

There are two Houses. Obviously, it’s pretty easy to imagine a tie. If there is a tie, meaning one House has voted one way and the other has voted a different way, then §15 declares that the slate of electors which has been “certified” by “the Executive” in the state — presumably the governor — is the slate that gets counted.

What if the tie is bogus? Meaning, what happens if the alternative slate is completely bogus?

This is where things get difficult. 

There’s no doubt that the ECA expects Congress to be applying a legal standard to the facts as they’ve been presented to them. That’s their duty — to act like a court. 

But Congress is not a court. And it’s incredibly unlikely that any court, including the Supreme Court, would ever tell Congress how it must vote. 

Here is an extreme case: Louisiana is a solidly Red state. President Trump is forecast to win that state by about 57% of the vote. But Louisiana has a Democratic governor. So imagine after the people have voted, the Democratic party insists that the election was rigged, or COVID blocked Democratic voters from participating, or whatever. And based on those arguments, Governor Edwards certified a Democratic slate of electors from Louisiana. Both slates, or groups, of electors gather on December 14 and vote. Both send their votes to Congress. On January 6, Congress gathers to count the electoral votes. When they get to the state of Louisiana, they discover two slates, one for Trump and one for Biden, but only the slate for Biden bears the Governor’s signature. Two members (one Senator, one Representative) object.

Assuming the Senate is still Republican in the new congress, and assuming Congress then behaves in a completely partisan way, we’d expect the Republican Senate to vote to support the Trump slate, while the Democratic House would vote to support the Biden slate. That means there would be a tie between the two. Under the rules of §15, it is the Biden slate that would prevail because it is the slate signed by the Governor.

Can Senate Republicans count the votes before January 3rd, when the new Senate is seated?

You have probably noticed by now that a few of the issues above arise only when the Republicans keep the Senate. For example, if Democrats win the Senate and keep the House, the tie between the two Houses, discussed above, is unlikely to occur. In such a blue wave case, the question is: Can Mitch McConnell decide to count the slates before the new Democrat Senate is seated on January 3? 

The simple answer is no: under current law, the votes must be counted on January 6. But some have argued that this law is unconstitutional because it infringes on the right of the current Congress to decide on procedural questions that are directly committed to it.

Even if Senator McConnell argues the ECA is unconstitutional, thereby invalidating its designation of January 6, he cannot unilaterally reset the date of the joint session.  From the election of 1792 through the election of 1932, U.S. law set the date of the joint session as the second Wednesday in February; this was back when the President was inaugurated in March.  But when the 20th amendment moved inauguration from March to January (to reduce the lame duck period), Congress updated the law with a new date for the vote count.  U.S. law designated January 6 beginning with the election of 1936.  It’s necessary for Congress to choose a date by law because the Constitution does not set a date, nor does it grant the President of the Senate the power to do so.  Any joint session must be scheduled by law, including the cooperation of the House of Representatives.  

McConnell would create a giant trap for himself if his caucus refused to participate in the January 6 counting session.  Because any makeup session would require a new law, with the consent of the House, Nancy Pelosi could just refuse to reset the date.  And if the joint session doesn’t meet before January 20 . . . well, Nancy Pelosi becomes Acting President.  So it’s actually Nancy Pelosi, not Mitch McConnell, who may have the most to gain from arguing the ECA’s choice of January 6 is unconstitutional.

Furthermore, it’s unlikely that any joint session would count unless it meets the constitutional requirements of a quorum.  The constitution mentions two different types of quorums:  the first is a majority of members (of either house) to conduct business; the second is the quorum for the failsafe process to select the President and Vice President, should the electoral college count produce no winner.  That second quorum requires a House quorum of at least one representative from two-thirds of the states, and the Senate quorum requires two-thirds of members present.  As currently constituted, both the Democrats and the Republicans have enough members to block a joint session quorum under either standard.  No joint session can meet with a quorum without bipartisan support.

What happens if the winning candidate passes away after November 3rd?

Understanding that this could always be a possibility, some States have drafted their pledge laws to give electors voting discretion when their candidate has died. See, e.g., Cal. Elec. Code Ann. §6906; Ind. Code §3–10–4–1.7. According to Chiafalo, it is suspected that in such a case, States without a specific provision would also release electors from their pledge. Still, it’s noted that because the situation has not been before us, nothing in this opinion should be taken to permit the States to bind electors to a deceased candidate. In short, for the states that have not passed such a pledge release provision, the situation is unclear.

Many states have laws that, if read literally, would force presidential electors to cast votes for candidates who have won the state’s popular vote — even if the candidate were deceased. Take Colorado, whose law was before the Supreme Court. Colorado law says “each presidential elector shall vote for the presidential candidate . . . who received the highest number of votes” in the general election.

The Supreme Court upheld this law as valid, which means that, in the usual course, state officials can — indeed, must — reject electoral votes cast for any presidential candidate who did not win the state’s popular vote. That means only Trump or Biden will be eligible for electoral college votes in Colorado this year, with the winner of the state’s election receiving Colorado’s nine electoral votes.

It should also be noted that §1 of the  20th Amendment to the Constitution states: “The terms of the President and Vice President shall end at noon on the 20th day of January, and the terms of Senators and Representatives at noon on the 3d day of January, of the years in which such terms would have ended if this article had not been ratified; and the terms of their successors shall then begin.” 

§3: “If, at the time fixed for the beginning of the term of the President, the President elect shall have died, the Vice President elect shall become President. If a President shall not have been chosen before the time fixed for the beginning of his term, or if the President elect shall have failed to qualify, then the Vice President elect shall act as President until a President shall have qualified; and the Congress may by law provide for the case wherein neither a President elect nor a Vice President elect shall have qualified, declaring who shall then act as President, or the manner in which one who is to act shall be selected, and such person shall act accordingly until a President or Vice President shall have qualified.”

What happens if the process for determining the winner of the presidential election is not completed by January 20, 2021?

The President’s current term of office ends on January 20, 2021 no matter what. Under no circumstances may the President’s current term be extended. An incumbent president may remain in office after the expiration of his first term only by winning re-election for a second term.

In every previous presidential election, the winner was determined by the time the prior term expired. But the Constitution has a mechanism to deal with a situation where that’s not true. If, for whatever reason, the process for selecting the President has not concluded by the beginning of the next term, then the Twentieth Amendment provides that the Vice President shall “act as President” until a new President is chosen. If neither the President or the Vice President has been chosen by January 20–which could happen because the presidential electors vote for a ticket of a President and a Vice President together–then the next in line to “act as President” is the Speaker of the House.

What is the role of the Supreme Court in deciding the presidential election?

The most famous example of the Supreme Court intervening in a contested election was the Court’s decision in Bush v. Gore during the 2000 election. That case focused on the complexities of the recount in Florida’s very close race between Bush and Gore. After a chaotic and uneven recount process, the Supreme Court of Florida ordered a statewide recount of all votes cast in the presidential election. The Bush campaign challenged the ordered recount, alleging that Florida’s recount measure violated the Equal Protection Clause because the state did not have a uniform recount mechanism (meaning that each county within the state had different means for undertaking the recount). In a 7-2 decision, the Court found that there was indeed an Equal Protection violation. However, even in the face of an Equal Protection violation,  there was still an open question: whether an alternative recount method could be used to remedy the violation. In a 5-4 decision, the Court’s conservative justices found that no alternative recount mechanism could be implemented to satisfy the safe harbor deadline imposed by the 3 U.S.C. § 5. The Court’s decision ultimately halted the recount and thus effectively handed the election to Bush. 

There are a number of legal issues discussed in this FAQ that could theoretically end-up before the Court. Essentially, if there is a dispute over the meaning/application of the Electoral Count Act or if there is some constitutional violation stemming from any of the above scenarios (for example, a Due Process violation in the situation where a state cancels an election), the Court would have jurisdiction to hear the question. Additionally, it is not inconceivable that an aggrieved party may attempt to argue that the Electoral Count Act itself is unconstitutional — a claim that the Court would certainly have the power to hear. 

There is an argument, however, that concerns justiciability: namely, that the Court does not have the jurisdiction to hear issues related to a presidential election, as those issues should be left to Congress. Within constitutional law, there is a theory known as the “political question doctrine” that states that there are certain issues that are inappropriate for judicial review and should be left to the politically accountable branches. In the 1972 case O’Brien v. Brown, the Court itself seemed to signal that elections fall under the political question doctrine. The O’Brien Court explicitly noted that there is a “large public interest in allowing the political process to function free from judicial supervision” in questions involving the electoral process. Based on this theory, some scholars have argued that the Court did not have the authority to decide cases such as Bush v. Gore. It is not inconceivable to imagine that the Court would rely on the political question doctrine to avoid intervening in a future election dispute. The Court may want to avoid the kind of controversy that was created in the wake of the Bush v. Gore decision and simply take itself out of a political fight based on justiciability concerns.

How can lower courts remedy real or claimed election irregularities?

Lower courts derive their “remedial powers”—the ability to right wrongs—from 3 sources: federal laws protecting federal rights, statewide election laws, and powers inherent in the courts called “equitable powers.”

Statewide election laws deal largely with procedural matters of how the votes are counted or a result is challenged. However, if there are suspicions of fraud or just widespread technical deficiencies in ballots, courts will likely figure out an appropriate remedy using these equitable powers, and looking to past court decisions for inspiration. If courts have to decide how to provide a remedy in a presidential election, they will likely look to their past election dispute cases, and consider if the normal remedy is appropriate for a presidential election’s unique circumstances. The usual remedies vary by state, sometimes significantly. 

Strict Treatment

Florida is known for strict treatment of absentee ballots in local or statewide races. In the presence of widespread irregularities in absentee ballots, Florida courts have thrown out all absentee ballots and picked the winner based on the remaining votes, as happened in the 1997 Miami mayoral election. However, given the importance of a presidential election and the resources available to determine the true winner, this remedy would be inappropriate so Florida courts would likely hold an exacting recount instead. 

North Carolina courts have been similarly strict. In one election, a failure of voting equipment resulted in a failure to count a substantial number of absentee ballots. The dispute over that election was ultimately resolved by the state legislature, in accordance with preexisting state law. That election could set an important precedent, because an analogous state law provides that the state legislature may appoint presidential electors if the outcome of the election is not otherwise determined by a certain date.  Additionally, in exceptional circumstances, where an unsuccessful candidate can demonstrate that they would have been successful in the absence of irregularities, North Carolina Courts have invalidated the defective election and ordered a new one.

Iowa recently amended its election code (presumably in anticipation of the 2020 election) to disallow election officials from supplementing missing or incorrect ballot information with information from the voter registration rolls. Previously, a ballot with an irregularity could be corrected by officials so long as voter intent could be ascertained. However, under the amended statute, and following a recent Iowa Supreme Court decision, any irregularity is fatal and cannot be corrected even when it is clear who the voter intended to cast their ballot for. The court’s rationale was that the ballots contain clear instructions and it is therefore not a burden on voters to disregard an irregular ballot.

Nevada courts are also apparently more willing than other states to throw out ballots that contain irregularities.  In one case, the Supreme Court of Nevada held that a ballot containing marks other than those required for voting would not be counted. The ballots that contained extraneous markings were accordingly thrown out.  This precedent could be problematic for claims of voter fraud, especially regarding mail-in ballots that are missing information or contain extraneous markings. As further evidence of this precedent, the Supreme Court of Nevada held that ballots marked in pencil, instead of with a rubber stamp as required by law, must be thrown out. Despite the lack of modern case law on the subject, it would seem that Nevada leans toward rejecting ballots with irregularities.

Arizona is similar, though not as strict. While Arizona also believes the procedural requirements surrounding absentee ballots are important and must be followed strictly, it does not invalidate absentee ballots as a class, but does tend to require new elections when irregular absentee ballots either changed the election or cause reasonable doubt about the outcome, as was the case when, for example, a county official failed to match signatures on ballots to signatures on a voter registration list.

Lenient Treatment

In contrast, Ohio has a lenient policy towards irregularities, attempting to count absentee ballots when the will of the voter can be reasonably ascertained. For example, when ballots cast by voters but mailed in by a 3rd party decided the outcome of a school board election, the Supreme Court of Ohio upheld the election results. Ohio courts tend to rely on statutory remedies more than equitable remedies.

Georgia also treats ballots liberally. For example, in one case, a court found that irregularities with birthdates on absentee ballots weighed in favor of a preliminary injunction so issues with birthdates could be corrected so those ballots could be counted. The harm to voters not being able to cast their votes is considered a difficult standard to overcome. Indeed, in the case above, Georgia argued the state had a significant interest in protecting against voter fraud, meaning that absentee ballots with insignificant irregularities (like an omitted birthdate), should be counted.

Wisconsin has a similar statutory mandate to “give effect to the will of the voters, if that can be ascertained from the proceedings. . .” This has been generally interpreted to uphold election results when there has been substantial compliance with their terms. However, where substantial irregularities occur, Wisconsin courts have ordered remedies setting aside the outcome of an election.

Minnesota’s election statutes provide an extensive list of potential irregularities along with guidance for how to treat them. The guiding principle in these statutes is that every vote should be counted as long as it is possible to ascertain the voter’s intent. Thus the statute, echoed by the Minnesota Supreme Court, provides that misspellings, missing or incorrect birthdates, or other markings will not prevent a ballot from being counted.

Pennsylvania has a fairly rich body of precedent affirming the state’s reluctance to set aside an election. In general, the state would rather remove any fraudulent ballots from the total count while retaining the remaining votes. In cases where it is not possible to ascertain which specific ballots are fraudulent, the Pennsylvania court will not consider that to be sufficient to set aside an election; only ballots that are proven to be faulty can be purged. When faced with a ballot irregularity, the court has directed that officials must award those ballots to the intended recipient if that recipient can be determined by “clear and convincing evidence.”

A case from the Supreme Court of Nebraska is one of the only cases dealing with voter fraud in the state.  In a hotly contested city election, city officials lied to voters in order to gain their votes. The court held that although the statements were fraudulent, they did not invalidate the election. Based on this precedent, it seems that an allegation of fraud may be a difficult claim to make in Nebraska.

Somewhere in the Middle

Michigan has a policy that is in the middle. Michigan requires that challengers prove enough ballots were tainted that the election is cast in doubt, with “enough” typically interpreted as within the margin of decision or something close. The challenger does not need to prove that the tainted ballots actually voted in favor of the result. Michigan does not discard any class of ballots in modern times and instead tends to order new elections. Michigan courts tend to rely on a mix of equitable and statutory remedies, and the interaction between them can become quite complex.

In Texas, the outcome of an election is considered “materially affected” by Election Code violations. A violation (like the absence of a signature, notary stamp, or a late ballot) would require that outcome be set aside when a different and correct result would have been reached in the absence of the irregularities. However, only in clear cases of fraud will election results be overturned.  In cases of clear fraud or a violation of the Election Code, Texas throws out all of the affected ballots. But there is a high burden of proof to prove voter fraud, and the right to vote is zealously protected, so it seems an allegation of voter fraud would be more difficult to make in Texas than in states with strict treatment of ballot irregularities or fraud.

What happens if the candidates win an equal number of electoral votes?

With 538 electors it is possible that Joe Biden and Donald Trump might each end up with 269 electoral votes. What happens then?

Let’s begin with the Twelfth Amendment text specifying the requirement for a candidate to be elected president by the Electoral College:

the person having the greatest number of votes for President, shall be the President, if such number be a majority of the whole number of electors appointed;

The same rule applies to election of the vice president by the Electoral College.

Notice that a winner doesn’t simply need to have a majority of the electoral votes cast. The winner must have the electoral votes of a majority of the electors appointed. That means that if all 538 electors who can be appointed are, in fact, appointed, then 270 electoral votes are needed for election by the Electoral College, even if some number of these electors are deemed to have cast their electoral votes invalidly. Suppose, for example, that seventeen electors are deemed to have been appointed validly but not to have cast valid electoral votes, leaving only 521 electoral votes cast. Although 261 is a majority of 521 votes cast, 261 would not be a majority of the 538 electors appointed.

At this point the rules for electing the president diverge from the rules for electing the vice president. The Twelfth Amendment specifies the rules for how these contingent elections work if no one has received the electoral votes of a majority of electors appointed. Here is the text regarding election of the president:

…if no person have such majority, then from the persons having the highest numbers not exceeding three on the list of those voted for as President, the House of Representatives shall choose immediately, by ballot, the President. But in choosing the President, the votes shall be taken by states, the representation from each state having one vote; a quorum for this purpose shall consist of a member or members from two-thirds of the states, and a majority of all the states shall be necessary to a choice.

Notice that the House votes by delegation rather than the per capita. When the Committee on Postponed Matters originally reported the Electoral College scheme to the Convention the contingent election of the president took place in the Senate, with each member having one vote. Concerned that the Senate would have too much power the Convention shifted the contingent election to the House but preserved equal suffrage among the states. This was the Constitutional Convention’s concession to the small states. After being at a disadvantage in the Electoral College the small states would hold the upper hand in the contingent election.

A Contingent Presidential Election in 2020

In the current, 116th Congress Democrats hold about a 35 seat margin but Republicans have a majority in 26 of the delegation. If the presidential election is so close that it goes to the House the Democrats will probably have a reduced per capita majority and the Republicans will probably control a majority of the delegations.

No one expects anyone other Donald Trump or Joe Biden to receive any electoral votes in this election. If that turns out to be true then the House of Representatives can only choose one of those two as president. Constitutionally, there are no other options (unless one of them dies and Congress has never exercised its constitutional power to specify what to do in that case).

If Biden and Trump are the only two options nothing in the Constitution binds a member of the House to vote as her state voted or as his district voted. The Constitution leaves this election to a political body and we must expect that its members will exercise political judgment when deciding how to cast their votes.

No one has had a good word to say about the House contingent election for about two centuries. It was always on the chopping block in the first half of the nineteenth century whenever Congress considered revising the mechanisms by which we elect a president. Many proposals would have given the contingent election to a joint session of Congress with each member having one vote. Others would have sent the election back to the People. At least we can thank the Twentieth Amendment (ratified 1933) for putting the start of the new presidential term seventeen days after the start of the new congressional term so the newly elected House votes in the contingent election.

While the Convention shifted the contingent election of the president from the Senate to the House, it left the contingent election of the vice president with the Senate (whose President the vice president becomes.) Here is the text regarding election of the vice president.

… if no person have a majority, then from the two highest numbers on the list, the Senate shall choose the Vice-President; a quorum for the purpose shall consist of two-thirds of the whole number of Senators, and a majority of the whole number shall be necessary to a choice.

Two presidential elections have gone to the House and one vice presidential election has gone to the Senate.

The Contingent Election Before Ratification of the Twelfth Amendment

Prior to ratification of the Twelfth Amendment (1804) the electors cast two, undifferentiated electoral votes. The first place finisher became president if (1) he had the electoral votes of a majority of the electors appointed and (2) no one else had the same number of electoral votes. One a president was selected the candidate remaining with the most electoral votes was elected vice president unless there was a tie. (A majority was not needed.) If there was a tie the Senate broke it.

Although parties and tickets had emerged by the third presidential election in 1796 the electors were unable to distinguish their choice for president from their choice for vice president (even though most state statutes descried their role as electing a president and a vice president). In 1796 presidential candidate John Adams finished first with 71 electoral votes (out of 138 electors). Presidential candidate Thomas Jefferson finished second with 68 electoral votes. (One Maryland elector voted for both Adams and Jefferson.) Adams’ running mate Thomas Pinckney received 59 electoral votes. Jefferson’s running mate Aaron Burr received 30. Nine others received a total of 48 electoral votes including George Washington who received two.

If all of these anomalous electoral votes cast in 1796 resulted in a farce, the lack of them resulted in a tragedy in 1800. Once again President Adams ran with a Pinckney from South Carolina. This time it was Thomas’s cousin, General Charles Cotesworth Pinckney. Once again Thomas Jefferson ran with Aaron Burr.

The Adams-Pinckney ticket won 65 electors. The Jefferson-Burr ticket won 73. One of the Adams-Pinckney electors in Rhode Island sloughed off his electoral vote for Pinckney by voting for New York Governor and former Chief Justice John Jay so that Adams would have one more vote than Pinckney. One of the Jefferson-Burr electors somewhere should have done the same so that they would not end up tied. None did.

That left Jefferson and Burr in a first place tie, each with 73 electoral votes, a majority of the number of electors appointed. That sent their names – and no one else’s – to the House of Representatives controlled by John Adams party on a per capita basis and with neither Adams’ party nor Jefferson’s party controlling a majority of the 16 delegations.

The House held the contingent election immediately after Vice President Jefferson announced the results of the electoral vote on February 11. The first ballot failed to produce the necessary nine votes for anyone. Jefferson received the votes of eight states, Adams six, and Maryland and Vermont divided equally. The same result ensured on the second, third, . . . thirty-fifth ballots as February 11 turned into February 16. Historians now believe that some deal was struck so and that on February 17 the representatives from Maryland and Vermont who had voted for Burr cast blank ballots putting those states in Jefferson’s column and electing him president just 15 days before the start of the new presidential term.

Contingent Elections After Ratification of the Twelfth Amendment

Concerned that this episode not be repeated Congress sent the Twelfth Amendment, the Designation amendment, to the states in 1803 and they ratified it in time for the 1804 election. Following ratification the presidential election and the vice presidential election became concurrent processes rather than sequential processes.

One presidential election has gone to the House under the Twelfth Amendment. A different vice presidential election has gone to the Senate under the Twelfth Amendment.

The Contingent Presidential Election of 1824-1825

CandidateElectoral Votes 
Andrew Jackson99Retired General, Former Senator
John Quincy Adams84Secretary of State
William Crawford41Secretary of the Treasury
Henry Clay37Speaker of the House
Total261 

Andrew Jackson captured the most popular votes and the most electoral votes, but without the votes of a majority of the electors appointed the election went to the House of Representatives which could only choose one of the top three. Fourth place finisher Henry Clay presided over the election in which thirteen of the 24 state delegations voted for Adams on the first ballot. Andrew Jackson’s supporters alleged a corrupt bargain between Adams and Clay who would quickly become Adams’s Secretary of State. It would be of no avail. Jackson would have to wait four years before getting his revenge with a sound thrashing of Adams in the 1828 election.

Interestingly, there was no need for a contingent vice presidential election in 1824/1825. After initially entering the presidential race Secretary of War John C. Calhoun shifted his attention to the vice presidency. Running a campaign more or less independent of any of the presidential candidates Calhoun captured 182 electoral votes, far more than the 131 needed for a majority. Four years later Calhoun would be elected Jackson’s first vice president.

The Contingent Vice Presidential Election of 1836-1837

Unhappy with his insignificant role as Jackson’s vice president Calhoun did not seek reelection in 1832. In fact, he resigned the vice presidency to accept an appointment as one of South Carolina’s senators. Jackson ally Martin Van Buren ran as Jackson’s running mate in 1832 and four years later Van Buren ran for the presidency with Kentucky Representative Richard Mentor Johnson as his running mate.

Johnson had accumulated a lot of political enemies over the years and that mattered in the vice presidential election of 1836. Although 170 of 294 electors appointed vote for Van Buren for president, not all of them voted for Johnson for vice president. Upset that a Virginian had not been picked to be Van Buren’s running mate the Virginia Democratic convention nominated a slate of electors who pledged that they would vote for Van Buren but not for Johnson. After they were elected that is exactly what they did.

That left Johnson with 147 electoral votes, exactly half the number of electors appointed but not a majority. After the two houses of Congress assembled concurrently to count the electoral votes the Senate quickly elected Johnson vice president by a vote of 33-16.

The Key Difference Between a House Contingent Election and a Senate Contingent Election

Let’s return to the Twelfth Amendment texts governing the contingent elections.

the House of Representatives shall choose immediately, by ballot, the President. But in choosing the President, the votes shall be taken by states, the representation from each state having one vote

the Senate shall choose the Vice-President

The original text in Article II concerning the contingent presidential election is much the same. On the other hand, there is a slight difference in the Article II text concerning the contingent vice presidential election.

the Senate shall choose from them by ballot the Vice President

Here is how the Senate Journal records the vice presidential election in 1837.

The Secretary of the Senate shall call the names of Senators in alphabetical order; and each Senator will, when his name is called, name the person for whom he votes;

The Secretary having called the names of the Senators, respectively, in alphabetical order, the result was as follows:

The members of the Senate went on record with their votes.

That is not what happened in the House in 1825 (or 1801). In 1825 the House adopted the following rules.

5th. … A ballot box shall be provided for each state.

The Representatives of each state shall, in the first instance, ballot among themselves, in order to ascertain the vote of their state; and they may, if necessary, appoint tellers of their ballots.

Two days later the House followed this rule.

The members of the respective states having taken seats, as required in the 5th rule, adopted on the 7th instant, proceeded to ballot in the manner prescribed by the said rule; and the delegations of the respective states having placed duplicates of their votes in the two general ballot boxes, the said boxes were deposited on tables prepared for the purpose …

Whereupon The following named members were appointed by the states, respectively, tellers to count the ballots, and to report the result to the House;

The tellers proceeded to examine and count the ballots, and having completed the same, and the votes in the two boxes agreeing, the tellers reported that the votes of thirteen states had been given for John Quincy Adams, of Massachusetts; that the votes of seven states had been given for Andrew Jackson, of Tennessee: and that the votes of four states had been given for William H. Crawford, of Georgia:

The language in the House Journal for the one and only post-Twelfth Amendment House election of the president is clear. The members place their ballots into ballot boxes. (They could not place voice votes into a ballot boxes.). The states placed their resulting votes into different ballot boxes.. A summary of the vote by states was made public in the House Journal. But not the vote by state and certainly not the ballots of the members, although their attendance is a matter of public record appearing in the House Journal.

Power of State Legislatures in Presidential Elections

Early caselaw distinguishes three issues: qualification of voters (who can vote), election of state and local officials, and election of federal officials, including the electors for President and Vice President. [1] The issues should not be confused, as they come up in a variety of lawsuits involving issues that are similar on the surface but quite legally distinct. Electors for President and Vice President can be selected in a broad variety of ways by State Legislatures, including statewide election (the most common method currently in use), apportionment by district (as in Maine and Nebraska), or through direct appointment by the Legislature (which was used in the early years of the Republic). Article I, Section IV of the Constitution specifies that 

“[t]he times, places and manner of holding elections for Senators and Representatives, shall be prescribed in each state by the legislature thereof;” Article II, Section I, grants an even more open ended power: “[e]ach state shall appoint, in such manner as the Legislature thereof may direct, a number of electors, equal to the whole number of Senators and Representatives to which the State may be entitled in the Congress.” 

This affirmative grant of power through the Constitution overrides the usual function of state governments.  The state legislatures are granted broad discretion by the U.S. Constitution as to Presidential and Vice Presidential electors, whereas state legislatures are bound by their own State Constitutions on issues of voter qualification and state and local elections. [2] There are two views on the limits of this power, which we colloquially refer to as the “superpower theory” and the “super-duper power theory.” 

The “superpower theory”emphasizes that the U.S. Constitution grants broad discretion to legislatures, but that the act of selecting electors is a lawmaking act. Therefore, though the legislature is unconstrained by substantive requirements in their state constitution, they must obey certain procedural elements for lawmaking, because the act of choosing the method of elector appointments is an act of ordinary lawmaking. [3] Some versions of the theory even require presentment to the Governor. [4] This theory argues that “a legislature must function in the method prescribed by the State Constitution in directing the times, places, and manner of holding elections for senators and representatives in Congress, since in so doing it is exercising the function of lawmaking, it does not necessarily follow that when functioning in the manner prescribed by the State Constitution, the scope of its enactment on the indicated subjects is also limited by the provisions of the State Constitution.” Com. ex rel. Dummit v. O’Connell, 298 Ky. 44, 50 (KY App. 1944). This theory gained prominence shortly after WWI, and was applied to many WWII era cases. [5] 

The “super-duper power theory” emphasizes the Lockean conception of the legislature as the preeminent branch of government, and holds that the State Legislature exercises “plenary” power [6] in determining how Presidential and Vice Presidential are selected “untrammeled” by the State Constitution, [7] with no restrictions on substance except those present in the U.S. Constitution, and without having to abide by the normal lawmaking process. “‘Whatever provisions may be made by statute, or by the state constitution, to choose electors by the people, there is no doubt of the right of the legislature to resume the power at any time, for it can neither be taken away nor abdicated.’ Senate Rep. 1st Sess. 43d Cong. No. 395.” McPherson v. Blacker, 146 U.S. 1, 35, 13 S. Ct. 3, 10, 36 L. Ed. 869 (1892) (quoting a report by Senator Morton). This view has support from early cases, which intensified around the time of the Civil War.  There is also language in Chief Justice Rehnquist’s concurrence in Bush v. Gore that arguably hints at this view, describing the executive’s role in the election as a mere agent of the legislature. [8] 

In the week preceding the 2020 election, two Supreme Court decisions have been directly related to the question of the legislature’s power in a presidential election. These two decisions illustrate the existing divide in the Court on the powers of the legislature. In a statement by Justices Alito, Gorsuch and Thomas, they seem to suggest the “super duper power” theory is law. Specifically, the Justices indicate that the legislature can not be constrained by the Constitution of the state. They stated, “The Supreme Court of Pennsylvania has issued a decree that squarely alters an important statutory provision enacted by the Pennsylvania Legislature pursuant to its authority under the Constitution of the United States to make rules governing the conduct of elections for federal office.” Republican Party of Pennsylvania v. Boockvar, 20-542, 2020 WL 6304626, at *1 (U.S. Oct. 28, 2020). In a related decision, Justice Gorsuch, in a concurrence joined by Justice Kavanaugh, also seems to support the super duper power theory. He wrote, “The Constitution provides that state legislatures—not federal judges, not state judges, not state governors, not other state officials—bear primary responsibility for setting election rules.” Democratic Natl. Comm. v. Wisconsin State Legis., 20A66, 2020 WL 6275871, at *2 (U.S. Oct. 26, 2020). Yet, even amongst the conservative Justices, there seems to be controversy on this issue. Justice Roberts specifically distinguished Democratic Natl. Comm. v. Wisconsin State Legis from Republican Party of Pennsylvania v. Boockvar. He suggested that state legislatures are bound by state court’s decisions, which would undermine the “super duper power” theory. See Democratic Natl. Comm. v. Wisconsin State Legis., 20A66, 2020 WL 6275871, at *1 (U.S. Oct. 26, 2020).

Given the conflict in the court at the moment, it is important to consider what reading of the law is most consistent with other decisions of the court recently. Since there’s been few decisions in recent decades that directly speak to the question of the superpowers of the legislature in picking the presidential electors, it’s difficult to determine what interpretation should control. It is best to consider recently litigated related issues. One consideration is what “the Legislature” means in this context. The Supreme Court has decided that “Wherever the term ‘legislature’ is used in the Constitution, it is necessary to consider the nature of the particular action in view.” Smiley v. Holm, 285 U.S. 355, 366 (1932). Essentially, there are two ways to construe the term “Legislature” here. It can mean either, “the representative body which ma[kes] the laws of the people,” or it can mean “[t]he power that makes laws.” Arizona State Legis. v. Arizona Indep. Redistricting Commn., 576 U.S. 787, 813 (2015). The first definition only includes the legislative bodies of the states that are analogous to the House of Representatives and the Senate. For example, in Florida, that would only include the Florida House of Representatives and the Florida Senate. The second definition would include both of those bodies, as well as the other legal bodies that constrain them, such as the Florida judiciary.

Arizona State Legis. v. Arizona Indep. Redistricting Commn. (2015) is the most recent Supreme Court case to decide that “Legislature” could mean something more than the representative bodies. In that case, the people of Arizona argued that they were allowed to create an independent body for the purpose of redistricting legislative and congressional districts, to prevent gerrymandering. Since redistricting congressional districts falls within the state legislature’s power to control the “manner” of electing representatives, a power granted to the legislature under Art. I, § 4, cl 1., the state legislature argued that only they had that power. The Court ruled 5-4 that the people had that power because “Legislature” does not necessarily mean representative bodies. Given the parallel structures of Art. I, § 4, cl 1., and Art. II, § 1, cl 2., this decision supports the assertion that Art. II, § 1, cl 2, could be interpreted the same way.

However, Arizona State Legis. v. Arizona Indep. Redistricting Commn. (2015), was a 5-4 partisan decision in favor of the “liberal justices.” Three sitting justices at the time are no longer on the court (Scalia, Kennedy, and Ginsburg), and there are a total of three new Trump-appointed justices on the court (Gorsuch, Kavanaugh, and Barrett). Given this massive shift, a recent 5-4 decision may not have much strength. This is why Rucho v. Common Cause (2019) is important. There, the court decided that the federal judiciary was not the correct body to determine issues like partisan gerrymandering. They then spoke positively about state efforts to combat partisan gerrymandering. Specifically, they spoke highly of the Florida Court’s decision to strike down redistricting in violation of Florida’s Fair Districts Amendment. This affirms that state bodies do have some power over the manner of elections. Importantly, this decision was handed down by a 5-4 conservative leaning court. Thus, the conservatives have accepted that, at least in the Elections Clause, other bodies than “the legislature” have control over the “manner” of choosing.

Another recent case to consider is Chiafalo v. Washington. There, the court considered whether states have the power to bind their electors to the state popular vote. Most importantly for the present consideration, the Court repeatedly minimized the role of the legislature in choosing presidential electors, and emphasized the importance of the people of the state choosing the electors.  This decision was 9-0, and decided just this past July. The general tenor of the decision would seem to suggest that the legislature may not have so much power to throw out votes of the people.

[1] On qualifications: “It belongs to the legislative department of the government to decide who shall have the right of voting, what shall be the qualifications of voters, and when and where the qualified voters shall exercise the right, except so far as the legislative authority is limited and restrained by the fundamental law; and this general authority over the whole subject belongs to the legislature of this State, with such limitations only as are imposed by the Constitution of this State and that of the United States.” In re Opinions of Justices, 45 N.H. 595, 597 (N.H. 1864);. On state/local elections and the contrast with election of federal officials: “The bill then under consideration, as stated in the opinion, related to the election of State, county and town officers, and any general expressions used could not have been intended to have an application more extensive than the subject matter then under consideration. We find in that opinion no mention of the Constitution of the United States, whereas the question as to the election of Representatives to Congress and of Electors of President and Vice President is governed wholly by the Constitution of the United States as the paramount law, and the Constitution of this State has no concern with the question” Id. at 599 

[2] See footnote 1

[3] “In short, we think that the holdings of the Supreme Court in the case last referred to, so far as they are applicable to the question before us, may properly be said to mean no more than that the legislative process must be completed in the manner prescribed by the State Constitution in order to result in a valid enactment”

Com. ex rel. Dummit v. O’Connell, 298 Ky. 44, 50, 181 S.W.2d 691, 694 (KY App. 1944)

[4] “That practice is eloquent of the conviction of the people of the states, and of their representatives in state Legislatures and executive office, that in providing for congressional elections, and for the districts in which they were to be held, these Legislatures were exercising the lawmaking power and thus were subject, where the state Constitution so provided, to the veto of the Governor as a part of the legislative process.” Com. ex rel. Dummit v. O’Connell, 298 Ky. 44, 49, 181 S.W.2d 691, 694 (KY App. 1944). 

[5] Smiley v. Holm, Secretary of State of Minnesota, 285 U.S. 355 (1932) originated this line of cases. Some lawyers also argued that McPherson v.Blacker provided a similar limit with the language “legislative power is the supreme authority, except as limited by the constitution of the state.” McPherson v. Blacker, 146 U.S. 1, 25, (1892). However, the O’Connell court rejected this interpretation because the McPherson court also stated “[t]he clause under consideration does not read that the people or the citizens shall appoint, but that ‘each state shall;’ and if the words, ‘in such manner as the legislature thereof may direct,’ had been omitted, it would seem that the legislative power of appointment could not have been successfully questioned in the absence of any provision in the state constitution in that regard. Hence the insertion of those words, while operating as a limitation upon the state in respect of any attempt to circumscribe the legislative power, cannot be held to operate as a limitation on that power itself.”

McPherson v. Blacker, 146 U.S. 1, 25 (1892). Thus, Smiley is the proper originating case.

[6] Bush v. Gore, 531 U.S. 98, 104  (2000)

[7] “This provision of the Constitution, which gives the State legislature authority to prescribe the time, place and manner of holding elections, but reserves to Congress the power to make and alter such regulations, leaves no ground for serious doubt that the Constitution regards a law which prescribes the place where an elector shall vote, as a mere regulation for the exercise of his right, and not as a qualification of the elector within the meaning of the term as used in the Constitution; and therefore the legislature, under the Constitution of the United States which gives them authority to prescribe the place of holding elections for Representatives in Congress, exercise that authority untrammeled by the provision of the State constitution”

In re Opinions of Justices, 45 N.H. 595, 605 (N.H. 1864)

[8] “The state legislature has delegated to county canvassing boards the duties of administering elections.” Bush v. Gore, 531 U.S. 98, 116  (2000) (Rehnquist, C.J., Concurring). This arguably contradicts Smiley, but Justice Thomas has recently signaled he believes Smiley is limited to certain Congressional elections or redistricting, and thus would not apply to selection of Presidential and Vice Presidential electors.  Arizona v. Inter Tribal Council of Arizona, Inc., 570 U.S. 1, 34 (2013) (Thomas, J., Dissenting). Thomas’s argument, if accepted by the Court, may also limit the applicability of Rucho to the Presidential context.