Forthcoming article: "The Democracy Ratchet"

Over at SSRN, I've posted a draft of The Democracy Ratchet, forthcoming in the Indiana Law Journal. Comments welcome! The abstract:

Litigants seeking to lift burdens on the right to vote and judges adjudicating these claims have an unremarkable problem—what is the benchmark for measuring the nature of these burdens? Legal theories abound for claims under the constellation of rights known as the "right to vote." And when a legislature changes a voting practice or procedure, courts may have an easy benchmark—they can consider what the right to vote looked like before and after the enactment of the new law, and they can evaluate a litigant’s claim on that basis. Recently, federal courts have been relying on this benchmark for the principal causes of action litigants might raise after a new law has been enacted—a Section 2 challenge under the Voting Rights Act, a freedom of association claim subject to the Burdick balancing test, and an Equal Protection analysis derived from Bush v. Gore. And frequently, courts have found that new laws that eliminate once-available voting practices or procedures fail.

I describe this new practice as the Democracy Ratchet. But it is only recently that a convergence of factors have driven courts to (often unwittingly) adopt the Democracy Ratchet more broadly. So while a legislature can expand such opportunities, courts scrutinize cutbacks on such opportunities with deep skepticism—deeper than had no such opportunity ever existed. The ratchet tightens options, squeezing the discretion that legislatures once had.

This Article seeks to solve the puzzle of how courts have scrutinized, and should scrutinize, legislative changes to election laws. Part I identifies recent instances in which federal courts have invoked a version of the Democracy Ratchet. It identifies the salient traits of the Democracy Ratchet in these cases. Part II describes why the Democracy Ratchet has gained attention, primarily as a tactic of litigants and as a convenient benchmark in preliminary injunction cases. Part III examines of the history of the major federal causes of action concerning election administration—Section 2 of the Voting Rights Act, the Burdick balancing test, and the Equal Protection Clause. In each, it traces the path of the doctrine to a point where a version of the Democracy Ratchet might be incorporated into the test. It concludes that these causes of action do not include a substantive Democracy Ratchet. Part IV turns to determine how the Democracy Ratchet might be used. It concludes that the Democracy Ratchet is best identified as an evidentiary device and a readily-available remedy for courts fashioning relief. It then offers suggestions for its appropriate use. Part V identifies some concerns with existing use of the Democracy Ratchet and instances in which it may be incorrectly used. It offers guidance for courts handling changes to election laws. Part VI concludes.

Quick thoughts from today's oral argument in Minnesota Voters Alliance v. Mansky

Here are a few quick running thoughts from today's oral argument in Minnesota Voters Alliance v. Mansky. This post will be updated. The transcript PDF is here. (As an I aside, I wrote an article about the concept of "ballot speech," or the contents of the ballot itself as communicating expressive and informative content for voters, in this piece in the Arizona Law Review. Mansky involves a related question on restrictions on speech in the polling place.)

Express advocacy? Justice Kagan early in the argument, along with Justice Kennedy, wondered about more narrowly-drawn rules on matters like express advocacy for or against a particular candidate rather than broader political messages. That might be an attractive option for a Court looking to fashion a rule that offers the state some flexibility to regulate in the future.

Content and overbreadth: Justice Alito later pressed on this issue to wonder if candidate-based content might be somehow a separate matter properly subject to regulation. Justice Kagan in particular was concerned about how a proper overbreadth challenge might look. Justice Gorsuch later in the argument wondered about Minnesota acting as "outlier" when examining whether Minnesota had a compelling interest to justify the potential (as he said, "often undocumented") chilling effect.

Scope of intimidation: Chief Justice Roberts wondered about this notion of "decorum" in the polling place, emphasizing that freedom from intimidation is a distinct issue. At the same time, he wondered, "maybe bitter, sharp, political campaign going on, and maybe, just before you cast your vote, you should be able to have a time for some quiet reflection or to do that important civic obligation in peace and quiet without being bombarded by another campaign display." Later Justice Kennedy wondered about the difficulty of enforcing decorum if it largely turned on individualized determinations from polling officials.

Late in the oral argument, Justice Kagan wondered about how to evaluate "decorum." The courtroom was a good place for decorum, she thought. But she wondered why the polling place sounded "a little bit church-like," when it came at the end of "often a rowdy political process."

First Amendment issues at all? Justice Kennedy wondered, "Why should there be speech inside the election booth at all, or inside the what you call the election room?" From a justice usually known for his robust First Amendment views, this struck me as notable. Justice Ginsburg jump in to join the concern.

Facial challenge: One related question to the overbreadth concern was the scope of the challenge, as a facial challenge as opposed to as-applied. The Court's doctrine in this area has not been the most coherent, so I won't dig into issues now. But Chief Justice Roberts wondered about the "tiniest little logo" as being subject to the law and somehow affecting "decorum" as potentially a problem.

Arbitrary enforcement and defining political matters: Justice Alito wondered about the risk of arbitrary enforcement and the difficulty of election officials line-drawing in the application of this statute. In a series of hypotheticals testing this limit, Justice Alito got the state's attorney to say that a T-shirt with "the text of the Second Amendment' Could be viewed as political, but notthe text of the First Amendment. (Oral arg. transcript at 40.) It highlighted a very basic problem with a statute that had as broad a scope as Minnesota suggested--and perhaps suggests that the Court would require something narrower.

Justice Alito later worried about partisan election judges determining the political connotations of materials. The state's response? This is not terrible unusual, given that election judges make all kinds of determinations.

Burson: The Court showed no interest in overruling Burson. (But such things may remain unsaid....) Late in the oral argument, Justice Gorsuch seemed satisfied that Burson would be the narrower case of "campaign speech," compared to Minnesota's law of "additional political speech." But, returning to the definitional concerns of the Court noted earlier, that may not be satisfactory.

Compelling interests: Near the end, Chief Justice Roberts emphasized that it did not appear that the state's interest were "terribly strong." Only time will tell....

New essay draft: "Legal Quandaries in the Alabama Senate Election of 2017"

I have posted a new essay forthcoming in the Alabama Law Review, entitled Legal Quandaries in the Alabama Senate Election of 2017. Here is the abstract:

President Donald Trump’s decision to nominate Alabama Senator Jeff Sessions as his Attorney General resulted in a vacancy in the Senate and triggered a special election. The special election, however, revealed the many complexities of the Seventeenth Amendment, special elections generally, and Alabama state law specifically.

This Article traces a series of legal quandaries that arose from the special election, some of which remain open questions for future Alabama elections, and for United States Senate elections more generally. Part I examines the scope of the Alabama Governor’s power to call for a special election under the Seventeenth Amendment and state law. Part II scrutinizes the complications for replacing a late-withdrawing candidate and for counting votes cast for a candidate who resigns. Part III identifies proposed gambits, from postponing the election to write-in campaigns, that never came to fruition. Part IV examines the timing surrounding certification of election results in Alabama. Part V looks at gaps in Alabama’s recount and election contest procedures. Finally, Part VI identifies the most significant opportunities to clarify Alabama law and to properly interpret the Seventeenth Amendment to avoid uncertainty in future elections.

I have a very short turnaround before submitting the final draft for editing, but I welcome any comments or feedback!

A small data point on an emergency designation in a California election

Emergencies and elections don't get along well. The threat of emergencies, like acts of terrorism or massive weather events, remain a concern in elections, but we seem to lack many structures in place to handle such events.

Fires in Sonoma County, California prompted Governor Jerry Brown on October 19, 2017 to sign an executive order declaring an all mail ballot election for the November 7 election.

I was a bit skeptical of what I thought to be a fairly late emergency declaration and wondered how it might play out. Granted, it's an admittedly low-turnout election, but on the surface it appears that canceling in-person voting had a negligible impact, if any, on turnout.

The 2013 election had 6364 absentee and 1248 precinct ballots for 35.2% turnout. The 2015 election had 7003 absentee and 1235 precinct ballots for 33.0% turnout. Surely, a high percentage of voters already casting absentee ballots helps minimize any damage from canceling precinct voting. The final results this election were 6590 absentee for 31.2% turnout--a decline, but no bigger than the previous off-year-over-year decline.

I'm certainly not equipped to address matters like the competitiveness of the candidates or contentiousness of the issues or any of the many other confounding variables that could affect turnout. But, it's a small data point to consider in the larger scheme of thinking about how to handle emergencies and elections.

Sexual misconduct, Senate precedent, and the Expulsion Clause

I recently wrote a piece at Law & Liberty on the Senate's power to expel Roy Moore in the event he wins the Alabama Senate special election. He has been accused of sexual misconduct, and calls for him to resign have not been heeded. Some have wondered about whether the Senate might refuse to seat him; I explain in the piece why that is inconsistent with the constitutional understanding set forth in Powell v. McCormack.

But what about the power to expel a member once seated? The Constitution says this:

Each House may determine the Rules of its Proceedings, punish its Members for disorderly Behavior, and, with the Concurrence of two-thirds, expel a Member.

I explain in the piece that there are few contours to the power to expel, except that it must occur by a two-thirds vote. That's mostly a procedural check, and the substance has largely been left to the Senate to flesh out. (Professor Brian Kalt shares similar thoughts in a Wall Street Journal opinion.)

I identify a couple of precedents that would limit the Senate's power to expel Mr. Moore. First, the Senate has typically wondered whether it has the power to expel members for conduct that arose prior to the candidate taking office. (Some earlier debates actually focused on whether it had to arise during that existing term in office and not from a preceding term, but recent Senate investigations have moved away from that view.) Second, the Senate has generally refused to expel a member for conduct known to the voters at the time of the election, the notion being that it's not for the Senate to expel a member with such baggange sent by the voters.

The Senate has not expelled a member since the Civil War, but it has investigated members who resigned before facing an expulsion vote, like Senator Bob Packwood in 1995.

New allegations have arise in recent weeks. Representative John Conyers and Senator Al Franken have been accused of sexual misconduct, and campaigns arose to urge their resignations.

Mr. Conyers heeded such calls this week and resigned, rather than face an ethics investigation and, perhaps, subsequent expulsion vote. His conduct arose during his terms in office (as he has been in the House for several decades).

Mr. Franken is facing new calls from a number of his fellow Senators to resign. Some allegations of his sexual misconduct arose from events prior to taking office, but certainly were not widely known to voters. He is also facing an ethics investigation.

It is worth noting how the rather underdefined precedents shaping Congress's practices under the Expulsion Clause may be experiencing a rather significant change right before our eyes. Ethics investigations and calls for resignation for (at least some) conduct arising prior to taking office suggest that a good number of Senators would apply the Clause in a way that the Senate has been reluctant to do in the past.

This is not to say that this is a wrong view, just to say that the Senate is, as is within its authority, deciding the proper scope of its authority under the Clause. It may be using informal means, like social media campaigns, political calls for resignations, and political party threats (perhaps offering primary challengers, withholding campaign funding, or revoking committee assignments), in ways that do not impact the Expulsion Clause.

But in the event Mr. Moore is seated, and the Senate begins to look at these recent precedents involving allegations of sexual misconduct prior to a candidate taking office, it may be moving toward a new understanding of its own power under the Clause.

That may not be the case, of course. Expulsion is an understandably extraordinary remedy, requiring significant consensus, and we have seen the Senate bluster recently and fail to carry through. These other tools at the disposal of the parties and the Senate may make the Senate reluctant to carry through with expulsion of Mr. Franken (if he does not resign in the next several hours) or Mr. Moore (if he is elected). But it is worth considering how recent weeks may be shaping, before our very eyes, a renewed attention to the Senate reconsidering its past practices and reinvigorating the Expulsion Clause.

A critically important legal question looms over Alabama Senate race: what happens if Roy Moore withdraws?

Election law informs political strategy. The choices that parties, candidates, and voters make may change depending on the legal consequences of those actions.

A critically important legal questions looms over the Alabama Senate race. If embattled candidate Roy Moore withdraws, it may have one of two effects. There is uncertainty about which effect will take place. And it could entirely change the political strategy of Republicans.

Imagine for a moment, and it requires little imagination, that Mr. Moore remains in the race. Republican leaders have a choice: stick with Mr. Moore, or back a single write-in candidate. (A third might be to simply endorse Democratic candidate Doug Jones, but I'll assume Republicans want to keep the seat in the hands of someone who'll caucus with them.) There isn't much to think about in this scenario. It's purely a question of political strategy.

Imagine, instead, Mr. Moore withdraws from the race, or the party withdraws him as its nominee. (I should pause on the last point to note that I hardly know how the party has a mechanism to do so, and whether it can yank the rug out of its own nominee, apart from the fact that state law apparently authorizes it; I won't address how it might do so, or whether Mr. Moore might challenge it.)

What is the strategic decision from the Republican Party then? In my view, it all turns on what Alabama Code 17-6-21(c) means. And while the Secretary of State John Merrill has said one thing about what the law says, I'm not sure that's accurate. And if it isn't accurate, the strategy changes completely.

If Mr. Merrill's interpretation is correct: if an ineligible candidate (dead, disqualified, or withdrawn) receives the most votes, the election is declared null and void, and a new special election would need to be called. That is consistent with a long line of precedent in Alabama, as recently articulated in law as 2001.

In this scenario, the best Republican strategy is to "punch Moore." Consider the 2006 Florida race involving disgraced representative Mark Foley, where his name appeared on the ballot but actually stood for stand-in replacement Joe Negron, and where the slogan "punch Foley for Joe" instructed voters that voting for the disgraced candidate would be the best choice--because it wasn't for the disgraced candidate at all. The same held true in Missouri in 2000, where voters chose the deceased Mel Carnahan and knew that the governor would appoint a replacement (Mr. Carnahan's wife, he promised) before the special election to replace him.

A write-in candidacy is not as good a choice. That is, because some early votes have already been cast for Mr. Moore; finding a capable candidate is dicey; and getting write-ins generally is a challenge... why not reset the clock?

If my interpretation is right: HB 62 was enacted in 2014. It changed the rules for recounts for late-withdrawing candidates only. Its amended text, as I've emphasized, provides: "In the event that a candidate submits a notification of withdrawal after the applicable deadline, the name of the candidate shall remain on the ballot and the appropriate canvassing board may not certify any votes for the candidate." (Similar language is in (b) when a party withdraws a nominee.) That means, the second-best vote-getter actually wins--because there are no votes for the withdrawn candidate, because no votes for that candidate have been counted.

In this scenario, the best Republican strategy is to back a write-in candidate. Early votes for Mr. Moore have already been cast, which is a problem. But, if Mr. Jones would be the winner, their sole priority would be to back a write-in candidate who could win--not votes for Mr. Moore that would be thrown out.

And a strategy to vote for Moore would lead to a Jones win. It's exactly the reverse incentives.

The resolution of this question is of critical importance. Litigation assuredly will ensue: if Mr. Merrill's interpretation is heeded, Mr. Jones would surely sue to claim that he would win, and perhaps seek a declaration before the election. Voters would be confused, weakening Republican chances in the event they wait. Mr. Jones's supporters might be surprised if Mr. Merrill's interpretation carries the day and he is fighting an invisible candidate through December 12.

Regardless of which rule is the right one under Alabama law, it is critically important that this question be addressed sooner rather than later.

Sorting out the Alabama Senate election: could the governor reschedule the special election?

Whew. After musing about possibilities in the Alabama Senate special election, and digging deeper into vote-counting for a withdrawn candidate, a new scenario has popped up: the governor postponing the election. As reported in the New York Times:

State law gives the governor broad authority to set the date of special elections, and Ms. Ivey, who is a Republican, already rescheduled the Senate election once, after inheriting the governor’s office in April when her predecessor, Robert Bentley, resigned in a sex and corruption scandal. Ms. Ivey’s advisers have not ruled out exercising that power again, according to Republicans in touch with her camp, but she has signaled that she would like reassurances of support from the White House before taking such an aggressive step.

. . .

But there is no apparent precedent for rescheduling an election so close to the planned vote, Republicans acknowledged. In addition to state election laws, such an extreme step could also run afoul of federal voting rights law.

The thought goes, I suppose, that a delay would give extra time for Mr. Moore to withdraw (voluntarily or involuntarily), and such a withdrawal, which might occur outside of the 76-day withdrawal window, would permit the party to replace Mr. Moore's name on the printed ballot.

Could this actually be the case? I'm not so sure. (I'll only address here a proposal to postpone the election, not to cancel it, and it would not require any action on the part of the legislature. I won't address whether independent Alabama constitutional rights would prevent this action from the governor in a voting rights context. I'll also emphasize, a I did in my last post, that this appears quite speculative, as Governor Kay Ivey has indicated no intention of changing the election date. Finally, I'll only discuss Alabama state law. I won't discuss the possibility that this act might run afoul of federal law or the United States Constitution.)

First, Ms. Ivey "rescheduled" the special election after Governor Robert Bentley resigned. Mr. Bentley had previously scheduled the special election to coincide with the 2018 regularly-scheduled primary and general elections. Each did so ostensibly pursuant to Alabama Code 36-9-8: "Whenever a vacancy occurs in the office of senator of and from the State of Alabama in the Senate of the United States more than four months before a general election, the Governor of Alabama shall forthwith order an election to be held by the qualified electors of the state to elect a senator of and from the State of Alabama to the United States Senate for the unexpired term. If the vacancy occurs within four months of but more than 60 days before a general election, the vacancy shall be filled at that election. If the vacancy occurs within 60 days before a general election, the Governor shall order a special election to be held on the first Tuesday after the lapse of 60 days from and after the day on which the vacancy is known to the Governor, and the senator elected at such special election shall hold office for the unexpired term."

But Mr. Bentley faced a legal challenge to his scheduling of the special election. He claimed that he had authority to push the election back to the general election. Challengers noted that he must "forthwith order an election" if the vacancy occurred "more than four months before a general election," and that only in instances where the vacancy occurred between four months and 60 days of the next scheduled general election could the governor hold the special election concurrent with the general election. A memorandum from the Legislative Reference Service concurred.

These are interesting matters of statutory interpretation under state law. But if Mr. Bentley lacked the power in the first place (that is, his original writ of election was inconsistent with Alabama law), then Ms. Ivey's pronouncement for an earlier special election was not a "change" or a "reschedule" of the special election; it was the first valid writ of election issued. While the Times describes this as a "rescheduled" election, the legal effect matters. If Mr. Bentley's original election lacked the force of law, Ms. Ivey practically (but not legally) rescheduled the election.

So, is there anything that constrains the governor? At least some constraints appear in the statute and its interpretation (apparently flouted by Mr. Bentley). But consider other relevant portions of the messy Election Code.

Consider constraints on the governor's power under Alabama Code 36-9-9: "The Governor . . . must give notice of a special election to elect a senator for an unexpired term in the same manner and for the same time as is prescribed for special elections to fill a vacancy in the office of members of the House of Representatives in Congress."

That refers to Alabama Code 17-15-3: "All special elections provided for by this chapter are to be ordered by the Governor, who must issue writs of election, directed to the judge of probate of the counties in which such election is required to be held and must specify therein the district or county in which, and the day on which, such election is to be held; the cause and object of the same; the name of the person in whose office the vacancy has occurred and, in all cases in which a special election is directed in a district composed of more than one county, such election must be directed to be held on the same day in each county." And Alabama Code 17-15-4, "The Governor must give notice of any special election for representatives in Congress, or state officers, by proclamation." Finally, Alabama Code 17-15-7, "Special elections are to be held and conducted, the returns thereof made and certificates given, and, unless otherwise expressly provided, regulated in all respects by the provisions in relation to general elections."

It's very hard for me to read these statutes together to suggest that the Alabama legislature has empowered the governor to schedule and reschedule, unilaterally, without constraint, special elections. First, 36-9-8 already places some timing restraints on the governor. Second, portions of Title 17 limit the discretion given to the governor, as special election are largely designed to resemble general elections.

The Seventeenth Amendment includes this relationship between the governor and the legislature: "When vacancies happen in the representation of any state in the Senate, the executive authority of such state shall issue writs of election to fill such vacancies: Provided, that the legislature of any state may empower the executive thereof to make temporary appointments until the people fill the vacancies by election as the legislature may direct.

I think the best way of harmonizing these statutes, and the way that events have transpired so far, is that Ms. Ivey's scheduling of the special election was the first valid proclamation from the governor (and that Mr. Bentley's proclamation was void under state law); and that she lacks the power to change the date. But, others may differ... and I'm always happy to consider competing theories or a more complete understanding of state statutes.

Sorting out the Alabama Senate election possibilities: what happens to the votes of a withdrawn candidate?

I blogged earlier about some of the possibilities in the Alabama Senate election. One question that arose was, what would happen to the votes of a candidate who withdrew but whose name could not be replaced on the ballot? (Let me explain at the outset I view the likelihood of Roy Moore voluntarily withdrawing, or the Alabama Republican Party invoking its power to remove him, as quite unlikely at this point, but the possibility remains.)

I explained that I thought that the votes for that candidate would not count and the second-place finisher would win. But a commenter pointed me to a recent statement from Secretary of State John Merrill:

What happens if Moore is withdrawn as the nominee but still receives the most votes?

Merrill said the election would be null and void. The second-place finisher would not win.

It would then fall to the governor to call another special election.

. . .

But there might be some question about that scenario. John Bennett, spokesman for Merrill, said one interpretation of the law is that if Moore is no longer a valid candidate but receives the most votes, Jones would be declared the winner.

Bennett said the official position of the secretary of state's office is that the election would be null and void, as Merrill said.

Let me start with the text of Alabama Code 17-6-21(c), emphasis added: "The notification deadline for persons who do not wish to accept nomination in a primary election is 76 days before the date of the election. A person who does not wish to accept nomination in a second primary election shall submit the notification set forth in subsection (a) before the printing of absentee ballots. The notification deadline for persons who do not wish to accept nomination in a general election is 76 days before the date of the election. In the event that a candidate submits a notification of withdrawal after the applicable deadline, the name of the candidate shall remain on the ballot and the appropriate canvassing board may not certify any votes for the candidate." (Similar language is in (b) when a party withdraws a nominee.)

The text of that statute, in my view, anticipates that the votes for the withdrawn candidate simply do not count. That would mean that the second-best finisher wins the election.

To explain some of the distinction, this is a centuries-old dispute about what to do with the votes in such a case: the "American rule," or the "English rule."

Many jurisdictions, including Alabama for a time, would count the votes for a deceased, ineligible, or otherwise withdrawn candidate. In the event that candidate won, the election would be null and void, the office declared vacant, and a new election held. Consider State v. Stacy (Ala. 1955): "In the case before us, where it affirmatively appears that the appellant received only forty-nine of the total number of votes, but the deceased candidate received the other 1,590 ballots cast for this office, we think the better rule is that the election for this office be held null and void because of the disqualification (by death) of the winning candidate." The Court there explained that voters might well not "waste" their votes by voting for a dead candidate; they may well know that the election would be found null and void, and they might prefer a subsequent special election. Another good reasons for this approach, too--consider that in Stacy, the runner-up received just 49 votes among over 2000 ballots cast! While we may accept plurality winners, a winner with such a tiny percentage of the votes may strike the public as something less than legitimate. Accord Banks v. Zippert (Ala. 1985); Ala. A.G. Op. 2001-041.

That was, for instance, the case of Mel Carnahan in Missouri in 2000, who died the week before the election and received the most votes. That led to the office being declared vacant, the governor appointing a Senator, and a special election was later held.

If that's historically true in many places in the United States, and in Alabama, what's the alternative? The "English rule" permits disregarding the votes cast for a deceased, ineligible, or otherwise withdrawn candidate. The second-best vote total actually wins, because those ineligible votes are thrown out.

It appears that Alabama has adopted the "English rule" in the context of late withdrawing candidates. That is, when a candidate withdraws within the 76-day window, votes for that candidate are simply not totaled. It is as if voters have cast blank votes (or "undervotes") for the office.

The Alabama legislature in 2014, with unanimous support, passed HB 62, which added this language to 21(c):

View fullsize

Note that the textual addition of the phrase at the end. It gives a new legal effect to late withdrawal--that is, the canvassing board "may not certify any votes for the candidate." That means, that candidate cannot receive the most votes. And that means, the "American rule" is not at play.

In the event Mr. Moore withdraws, then, any votes for him are not certified. The vote totals from the canvassing board would reflect vote totals for Doug Jones and for any write-in ballots cast. Whichever candidate among those names receives the most votes would win.

Note, too, this only extends to late-withdrawn candidates. In the event a candidate, say, died before Election Day, the "American rule" would still apply--the office would be declared vacant and a new special election would need to be called.

I have tried in vain to find any meaningful legislative history behind the addition of 21(c), but I think this is the best interpretation--and one that Mr. Merrill's spokesman admits is a possible interpretation.

Sorting out the Alabama Senate election possibilities in light of Roy Moore

After recent scandalous news surrounding Alabama Senate candidate Roy Moore, some have wondered what might happen in Alabama. There are just two candidates on the ballot, Mr. Moore and Doug Jones.

Mr. Moore cannot be replaced. Alabama Code 17-6-21(c) requires any withdrawal to occur at least 76 days before Election Day; we're about a month out now. It's the same deadline for parties to replace their nominees under 17-13-23. It used to be 45 days until a series of 2014 amendments.

It's understandable that we have pretty early withdrawal deadlines. The dramatic expansion of early and absentee voting around the country, coupled with the long lead time to print ballots and then mail them, especially to overseas uniformed military personnel, has pushed these deadlines out.

But, withdrawing does have an impact, even if no can can replace him. That is, if Mr. Moore withdraws from the race, votes for him would not be counted: "In the event that a candidate submits a notification of withdrawal after the applicable deadline, the name of the candidate shall remain on the ballot and the appropriate canvassing board may not certify any votes for the candidate."

UPDATE: Per an exchange with @Taniel on Twitter, I construe this to mean that in the event Mr. Moore withdrew, but still received the most "votes," the second-best winner would actually win the race, because those "votes" for Mr. Moore would not be counted. It would be as if there were no candidate at all. It does not mean that there is now a vacancy in the office. I think that's the best reading of 21(c) (and this provision was added in 2014, with no meaningful legislative history that I can find). FURTHER UPDATE: It appears that Alabama in 2014 overrode several previous state supreme court holdings on this subject. Alabama had adhered to the "American rule," which would count votes for deceased, ineligible, or other withdrawn candidates, and in the event that candidate won, the office would be declared vacant and a new election held. The "English rule," a version of which Alabama adopted, disregards votes cast for such candidates. I have found no particular motivation for doing so (as of yet!).

UPDATE: There's a similar set of rules in 21(b) for instances in which the party withdraws its support for the nominee. It isn't clear to me from the Alabama Republican Party how it might go about doing that--in a sense, stripping the nomination from Mr. Moore to render all votes cast for him as null. But, it is also there in the statute.

Anyone may be a write-in candidate in Alabama. Alabama law does not require pre-election filing requirements for write-in candidates. Its write-in law hasn't been used much but was recently heavily revised in 2016. That means that anyone can start a campaign right now and run as a write-in. The last write-in to win a senate election? Lisa Murkowski in Alaska, 2010. Before that? Strom Thurmond in South Carolina, 1954.

Alabama's "sore loser" law does not prevent Luther Strange from running a write-in campaign. Alabama does have a "sore loser" law that prevents a candidate from appearing on the general election ballot if he lost the primary election. But that condition does not apply to write-in candidates--only candidates whose names might appear on the ballot. The Alabama Secretary of State has made this clear. Some have mentioned memoranda from Mr. Strange expounding upon the "sore loser" law. But those refer to printing the name on the ballot; they do not extend to write-in candidates.

A court could pull a "Torricelli" from 2002. In 2002, Robert Torricelli withdrew 35 days ahead of the election. State law prohibited filling vacancies within 51 days of the election. The New Jersey Supreme Court in Samson concluded that in order to effect the state legislature's desire that the popular will of the people control the outcome of the election, and given that it would be feasible to replace Mr. Torricelli's name on the ballot despite some absentee ballots already having been sent out, it would be acceptable to create a judicial exception to the statute and permit a replacement.

There are good reasons, I think, why these kinds of judicially-created post hoc exceptions are not terribly persuasive to me; Bob Levy in 2002 offers his scathing critique here. In contrast, Professor Rick Hasen offered a defense of such moves under the Democracy Canon, to extend deadlines to benefit voters.

Regardless of what one thinks is right, it's a claim that might be made to the Alabama courts. Even there, complications remain. As Professor Hasen notes, many overseas voters have already received ballots, and it might be a challenge to get them replacement ballots in time for the election.

In the event Mr. Moore withdraws, however, recall that votes for him do not count. A court balancing the equities (assuming it's persuaded to create a judicial exception to the replacement statute) might conclude that in the event the candidate has withdrawn, and the overseas voters' votes for that candidate might not count anyway, it may be the case that replacing a candidate would be acceptable.

If this occurred, then, I think, Alabama's "sore loser" law would still remain in effect, and Mr. Strange could not be the replacement. But I'm less confident of how this might work....

In the event I have updates, I'll revise this post as necessary.