The #Calexit ballot proposal is unserious in a number of ways

I had perhaps foolishly entertained the notion that the #Calexit "independent California" movement might be a serious attempt to declare independence from the United States, and I described the hurdles facing a prospective ballot measure. I've now read the actual proposal, and it's entirely unserious, in a number of different ways, worth noting now that the proposal has been approved for circulation. I would read the text of the proposal first, which is shockingly minimalist. It does two things: amends the California Constitution to eliminate the provisions that California is "inseparable" from the United States and that the Constitution is the "supreme law of the land"; and calls for a "plebiscite questions" in 2019 on independence, which, if favorable, would require the governor to apply to the United Nations.

There a lot--a lot--of problems with this proposal. To name a few:

First, the initiative repeals the constitutional provision that "The state of California is an inseparable part of the United States of America, and the United States Constitution is the supreme law of the law." Never mind the dozens of other places in the California Constitution that refer to the United States of America or the United States Constitution--for instance, requiring state legislators to be citizens of the United States (Article IV, section 2) and the Governor (Article 5, section 2).

Second, the initiative makes no attempt to call for a constitutional convention, which would be necessary for this and other reasons--for instance, no need to elect United States presidential electors, senators, and representatives.

These problems alone indicate that #CalExit is not a serious independence proposal. Proponents claim this would be a "first step," but it is hardly any step at all--except, I suppose, an opportunity for earned media to treat the movement seriously despite the lack of legal impact of the movement.

Third, the initiative provides, via statute, for a "plebiscite question" given to the voters with the condition that at least 50% of registered voters participate for it to pass. California Constitution Article II, Section 10(a) provides no minimum participatory threshold for an initiative to take effect.

Fourth, the initiative provides, via statute, for a "plebiscite question" given to the voters with the condition that at least 55% vote "Yes." California Constitution Article II, Section 10(a) requires only a "majority" to pass initiative statutes.

Fifth, ballot initiatives are now to occur on the first Tuesday following the first Monday in even-numbered years, Cal. Election Code 9016. The initiative provides for elections in March of odd-numbered years and does not amend or refer to 9016 to account for this change. (Indeed, the California legislature recently amended this provision to ensure that major initiatives and referenda would occur in instances with higher voter turnout, March of odd-numbered years being among the lowest of turnout.)

Sixth, the "plebiscite questions" likely exceeds the power of the people of California to behave via initiative. California courts regularly conclude that the initiative power, even under "the most liberal interpretation," limit the power to the adoption and rejection of "statutes." Such conclusions come from AFL v. Eu and many California state appellate rulings. Declaring that the governor must behave in a certain way--that is, petitioning the United Nations for statehood--is not really a legislative enactment embraced by statute.

Seventh, and relatedly, the people of California may lack the power to direct the Governor to do something. For instance, in Widders v. Furchtenicht (2008), a California appellate court concluded that the initiative power was inappropriate to direct the city council to exercise its "informed judgment" in promulgating laws about housing and retail stores.

It's possible that some would consider the amendment to be a "revision" of the state's constitution, but I highly doubt that removing some largely precatory language--part of which is already redundant of the federal Constitution, anyway--amounts to a revision.

It's also worth noting these problems are not necessarily fatal to the measure appearing on the ballot. California courts in particular are reluctant to engage in pre-enforcement challenges on the merits of a proposed initiative, in the hopes that such questions might be mooted if the initiative fails to get enough signatures or fails at the ballot box.

Some of the problems outlined above might be cured with statutory construction that would give effect to the challenged provisions. Others might simply be notes about the limited nature of this question left for subsequent legal amendments. But they are, I think, serious reasons why the #Calexit proposal is decided unserious.

In today's WSJ: "Faithless Electors: Now It’s Up to Congress"

In today's Wall Street Journal, I have an opinion piece entitled, "Faithless Electors: Now It's Up to Congress." It begins:

The 538 members of the Electoral College convened Monday and cast a majority of their votes for Donald Trump for president and Mike Pence for vice president. When Congress convenes on Jan. 6 to count the votes, it will mostly be a formality. But its decision to count or exclude the votes of some “faithless electors” will set a precedent for future elections.

And it concludes:

These are challenging questions that cannot be answered by a judge or a court. Only Congress decides what to count. And while it won’t change the outcome of this election, its decisions will affect how states handle faithless electors in the 2020 election and beyond.

The Electoral College won't stop Trump--but it may change how political parties pick electors in 2020

The presidential election is quickly approaching--on December 19, in state capitals around the country, presidential electors will assemble and vote overwhelmingly for Donald Trump and for Hillary Clinton. It is almost guaranteed that Mr. Trump will secure at least 270 electoral votes, the minimum necessary, and probably something close to the 306 electoral votes he is presumed to receive. And while there have been discussions among "Hamilton Electors" to vote for someone else, the Constitution is designed to thwart such conspiracies across states, and Mr. Trump's presumptive lead is all but insurmountable. (Please note that while the Electoral College can elect someone other than Mr. Trump, or can elect no one at all, I am simply describing, as the links above suggest, why it is, among other things, a lead that is "all but insurmountable.")

Last week, five presidential electors in three states have filed lawsuits in federal courts seeking to strike down laws that purport to bind electors to vote for the candidates they are pledged to support. They seek to ensure that they have the right to vote for whomever they wish, regardless of the candidate they are supposed to support, and to have a court conclude that such laws instituting criminal or civil penalties are unconstitutional. (For the record, I agree that such laws may be unconstitutional.) The litigation seeks to limit the scope of Ray v. Blair, the 1952 Supreme Court decision that affirmed the ability of states to impose (non-binding) pledges on presidential electors. (The enforceability of such pledges was left for another day.)

But litigation occurs in a particular context, and there are significant procedural problems to these claims. Those problems may prevent courts from reaching the merits of such claims. These claims all seriously struggle from the likely defense of laches, as the electors have brought claims just days before the Electoral College is scheduled to meet--when they have been nominees for many months, and when they knew they would be called to serve as of November 9. It is not immediately obvious that the state laws in California and Colorado empower state election officials to remove "faithless" electors from their offices, which suggests that abstention might apply, or simply the application of a canon of statutory interpretation that invokes the constitutional avoidance doctrine. The pleadings of some do not make it obvious that the electors intend to violate their pledge, only that they want to liberty to do so, which may (perhaps) lead to ripeness issues or even the failure to state a claim.

But even setting these procedural issues aside, the curious nature of these claims is where they have been filed. They are purporting to be an "anti-Trump" movement. But, these are Clinton electors in states carried by Mrs. Clinton! That is, their movement would undermine the Democratic candidate's ability to succeed in the Electoral College! And even total victory in these states would yield a grand total of zero Trump electors voting for someone other than Mr. Trump!

Now, I suppose there are two long-game purposes in this effort. The first is for these electors to force a kind of "national conversation" about the independent judgment of electors and to (quite publicly) encourage Trump electors to join them and vote for someone other than Mr. Trump. (Of course, they were already voting for someone other than Mr. Trump.) While freeing Clinton electors from their pledge has zero impact on the bottom line--if every Clinton elector voted for, say, John Kasich, then Trump still wins with 306 electoral votes--it could spark discussion with other electors.

The second is that a ruling in one court, perhaps appealed to a circuit court or even the Supreme Court, would have a ripple effect in other jurisdictions with Trump electors. Given the procedural hurdles already in place, it is unlikely that this could happen, but remains a possibility.

Of course, further buried within these electors' lawsuits is that they have largely been filed by former (and, perhaps, current) supporters of Bernie Sanders, some of whom before even Election Day expressed public displeasure at the prospect of casting votes for Mrs. Clinton. While the salutary effort is something in the vein of "anti-Trump," in reality it seems to be driven more by anti... well, Democratic establishment, at least for these particular Colorado and Washington electors.

(It's also worth noting that multiple Trump electors have expressed opposition to Mr. Trump--two electors, one from Georgia and another from Texas, intend to resign when the Electoral College meets by refusing to show; a third, from Texas, once supported Mr. Trump but has since written a piece published in the New York Times opposing him.)

So these lawsuits are not really designed to stop Mr. Trump from securing 306 electoral votes (or, really, the 270 electoral votes he needs to win). But it has created some rather curious alliances. For instance, the Republican elected officials called to defend the law in Colorado have come out quite strongly against the plaintiffs--that is, these Republicans are aggressively defending Mrs. Clinton's electoral vote total in Colorado. And the Colorado Republican Party has intervened in the case--and the Colorado Democratic Party has not.

That said, it is, I think, less curious than one might expect, at least the behavior of Republicans and when viewed through a (perhaps) Rawlsian framework. One might take the myopic view and claim that Colorado Republicans are trying to defend Mr. Trump's election, but that strikes me fairly unlikely--consider the two long-game purposes I enumerated above, which are exceedingly remote; and consider that the direct impact of the litigation would undermine Mrs. Clinton's position far more than Mr. Trump's.

Instead, consider what it would mean in a state--any state, regardless of your partisan preference--if you had fairly settled expectations of the roles of electors, and even a law that carried some generic threat against electors who acted against their pledge, and those settled expectations were called into question. As a member of a political party or a loyal partisan official, such a result would be fairly horrifying. After all, it would mean that your formerly-loyal slate of electors would now be open to influence; and even if your party's slate of electors did not win this particular election, it would also affect your slate of electors in future elections when you did win.

Despite the fact that electors may prefer independence (and that the Constitution, in my view, mandates it!), parties certainly do not prefer it. It is a reason they are empowered in most states to choose the slates of presidential electors. And it is a terrific loss of power if those electors now expect to act freely--indeed, so freely that they may undermine the party's nominee. The rational behavior of partisan officials, then, would be to defend such laws quite vigorously, regardless of partisan affiliation.

I expect, then, that this behavior of presidential electors will fairly significantly alter the behavior of political parties selecting slates of presidential electors in 2020, particularly if parties are worried that the legal pledges and settled expectations from previous elections have been called into grave doubt. Party reforms are some of the easiest reforms, because they require no new laws. But I would expect, at least in some jurisdictions, to see to following changes.

First, I would expect to see delays in the selection of slates of electors. Parties typically nominate slates of electors in the spring or summer, often before the parties' nominating conventions (and sometimes even before the parties' nominees are known). But in most states, such slates need not be submitted until just weeks before the November election date.

Second, parties are likely to engage in far greater vetting of such nominees. By postponing the selection process, parties might be more inclined to choose electors who have already gone on record expressing support of the party's presumptive nominee.

Third, parties might institute more control over who qualifies as electors for their party. They often include rules that one must be a member of that party, such as someone who voted in that party's primary. But they may require longer periods of party affiliationor greater demonstration of loyalty before qualifying as an elector.

Fourth, parties may defer to the presumptive nominee in selecting slates of electors. It's understandable why Bill Clinton was an elector in New York, of course! And greater control to candidates would ensure greater loyalty for nominees.

It is the case, I think, that these electors' efforts to sue to undo state pledges will likely fail; and that even in success the Electoral College will not meaningfully affect the settled expectations of the outcome of this election. But after all this, when the dust settles, I anticipate some significant change in behavior from political parties to fend off future efforts from electors to undermine their own preferences.

New effort (doomed to fail) calls for presidential electors to collectively exercise independent judgment

Presidential electors will meet in state capitals around the country on December 19. They’ll vote for the next president and vice-president. We assume most of them will vote for Donald Trump and Mike Pence. But a group of anti-Trump electors, mostly Democrats, have sought to form an alliance around a consensus candidate who is neither Trump nor Hillary Clinton. They remind us that electors are supposed to exercise “independent judgment.” They hope to collectively exercise independent judgment--something of an oxymoron.

But the Framers expressly designed the Electoral College to thwart such schemes. They’ve repeatedly failed in the past, and they’re all but doomed to fail this year.

The Constitution's design

During the federal convention of 1787, the Framers worried that selection of the president would be the subject of political “intrigue” or fall into the hands of a “cabal” of decision-makers. Alexander Hamilton explained in Federalist 68 that the Electoral College would avoid such “mischief.” If electors assembled in a single place, Hamilton noted, it would invite “heats and ferments,” “cabal, intrigue and corruption,” and a selection process gone wrong.

Instead, electors from each state would assemble in that state, separated from the electors of all the other states. They would meet on the same day across these states, “detached and divided” from another. Hamilton emphasized that that could not engage in any “combinations” that would affect their independent judgment. Electors would vote for a president and a vice president based upon the deliberations in their own states, not from any agreement among electors.

Early attempts for electors to conspire with one another across the states failed badly. As originally designed, electors voted for two candidates: the candidate with the most votes became president, and the candidate with the second-most votes became vice president. Federalists in 1796 wanted John Adams as president and Thomas Pinckney as vice president. Electors tried to conspire to ensure that Pinckney received fewer votes than Adams; otherwise, a tie would be sent to the House of Representatives. They also needed to ensure that both Adams and Pinckney secured more votes than rival Thomas Jefferson.

But too many Adams electors cast their second votes for someone other than Pinckney. In the end, Adams secured 71 electoral votes and Pinckney 59—but Jefferson received 68 electoral votes, good enough for second place and to serve as Adams’s vice president. Federalists had been thwarted by the decentralized design of the Electoral College.

The Twelfth Amendment permitted electors to designate which candidate would be the president and which would be the vice president. And no effort to thwart a candidate's election has succeeded since--in part because the system is designed to thwart such efforts.

Intrastate electoral independence

When electors exercise their independent judgment, they do so because of the deliberative process that occurs within their state and almost never collectively crosses state lines. In 1828, for instance, seven electors voted for William Smith as Andrew Jackson’s vice president instead of John Calhoun—all seven were in Georgia. Thirty electors in 1832 voted for William Wilkins as Andrew Jackson’s vice president instead of Martin Van Buren—all thirty were from Pennsylvania, as was Wilkins. And in 1836, twenty-three electors abstained from voting for vice president instead of supporting Democratic nominee Richard M. Johnson—all twenty-three were from Virginia.

Indeed, as long as electors are casting votes (many years ago, proposals for an "automatic" Electoral College were floated, eliminating the human element), it is good for electors to exercise independent judgment. In 1872, for instance, it was good that most Democratic electors voted for someone other than Democratic presidential nominee Horace Greeley, who died after Election Day. Or for eight electors to vote for someone other than William Howard Taft’s running mate James Sherman, who died a week before Election Day.

But exercising independent judgment as individual is quite different from conspiring collectively toward a common outcome, and particularly different from conspiring across state lines.

Political parties

It's true that we have something quite different than what the Framers anticipated in 1787 (but was quite well-established by 1804 when the Twelfth Amendment was ratified): the two-party political system that still dominates our election system. The rise of political parties created stability in the process—while electors could not conspire across states, their common partisan affiliation and the party’s selection of a nominee brought stability to the process across the country. Voters (or state legislatures selecting electors) knew well in advance that the electors would support a particular candidate--the candidate that party nominated. These were party loyalists.

While it would be essentially impossible to conspire during the meeting of the Electoral College, as a practical matter, partisan loyalties offered contrasting visions for presidential electors, and the Electoral College quickly became a fairly stable and routine selection process between the candidates of two parties. Indeed, such loyalty became so obvious that today almost all states have stopped listing the names of electors on the ballot, listing on the electors.

This description provides two important conclusions. First, the ex ante nature of presidential electors' loyalties makes for fairly easy affiliation with a single presidential candidate. It has been an impossible effort to corral presidential elector support across the states ex post, sometime after Election Day but before the meeting of the electors.

Second, the two-party system did change how elections occurred--we ended the expectation that races would be resolved in Congress. Since 1804, just two presidential and one vice-presidential election have been resolved in Congress--the election of 1824, where four candidates secured electoral votes but no one secured a majority; and the election of 1836, where just enough Virginia electors cast votes for someone other than the presumptive vice president that the election was sent to the Senate (which voted for the presumptive vice president anyway). As originally designed, the thought was that independent judgment would rarely result in a majority, sending the election to the House--a notion that collective deliberation would not occur!

An contingent election in the House

Furthermore, it's worth emphasizing that this Electoral College effort will not send the presidency to Hillary Clinton. It is, at best, designed to turn at least 38 Republican electors (and perhaps some Democratic electors) to vote for someone else (perhaps John Kasich), depriving Mr. Trump of at least 270 electoral votes and sending the election to the House of Representatives. But, as I've noted before, if the Republican-led House and the Republican-controlled state delegations--led by individuals like Paul Ryan--did not stop Mr. Trump at much easier points early in this campaign (such as during the Republican National Convention)--I find it hard to believe it would choose to deny him the presidency at this moment. Again, while it would be within the House's prerogative to select among the top three vote-getters in the implausible event no one secured 270 votes on December 19, it is yet another unlikely result.

Collective action

Finally, it takes only a moment to recognize the massive collective action problem, built into the design of the Electoral College. What assurance to electors have that their counterparts in the other 49 states (and the District of Columbia) will act as promised? Even if many did agree in advance, it is quite another to trust that such decisions are being made elsewhere.

And there is a potential unraveling problem in the digital age--while conspiracies might have been impossible in 1787, they face the unraveling of a decision-making process across time zones. Electors typically meet at noon in state capitals--noon, local time. (A few around 1 pm, and perhaps others scattered around these hours.) Early-voting electors have no guarantees that later-voting electors are voting as promised; and later-voting electors can observe if early-voting electors defected, which increases the likelihood of their own defection.

An effort doomed to fail

This piece, I hope, describes why such an effort is doomed to fail. It might be the case that, as a normative matter, we would prefer electors to conspire across state lines. But the system is designed to thwart such efforts--and quite successfully. We have had 53 presidential elections since the passage of the Twelfth Amendment; the outcomes have never been altered by "faithless" electors, and only once (the vice presidential election of 1836) was the race sent to Congress, which resolved it as would have been expected from popular voting, anyway.

It might be the case, as many have suddenly discovered, that the Framers had wisdom in authorizing the independent discretion of electors. But it is also the case that the Framers decidedly created a system that would be built upon independence during the meeting of electors and thwart conspiracies among electors--perhaps another element of our constitutional design that could inform what it is likely to occur this December 19.

The Vermont Supreme Court oral argument in Cruz and Rubio "natural born citizen" litigation

You may have long forgotten about the "natural born citizen" litigation in the 2016 presidential election, which I chronicled this spring. One can be forgiven, because, as hot as the topic was, Ted Cruz dropped out of the race, most cases disappeared, and little has been thought about the matter (much less about the even more tenuous and rare litigation surrounding Marco Rubio). My forthcoming Fordham Law Review piece chronicles the procedural paths these cases took, urging courts to exercise caution before needlessly plunging into disputes that the political process could readily solve, or areas where the state legislature failed to give them express jurisdiction.

Procedure took center stage in Vermont.

The plaintiff, H. Brooke Paige, was one of the first to file a claim challenging Messrs. Cruz's and Rubio's status as "natural born citizens." Mr. Paige filed pro se and sued both candidates and the state of Vermont. The case was dismissed. Mr. Paige appealed, again pro se. And the Vermont Supreme Court scheduled 30 minutes of oral argument for November 30, 2016. Messrs. Cruz and Rubio waived the opportunity to appear at oral argument. The oral argument has been uploaded.

The Vermont Supreme Court did care about procedural matters and asked Mr. Paige almost exclusively about those procedural points. First, as neither Messrs. Cruz nor Rubio were on the general election ballot, one justice wondered whether the dispute was moot--but Mr. Paige emphasized they were on the primary ballot. (Later analogies to abortion mootness exceptions arose.)

Another justice asked that the Constitution has a requirement as to who can serve as president; but does that extend the same as to who can run for president? Mr. Paige identified them as the same standard.

As to standing, Mr. Paige claimed he sought ballot access in Vermont.

The government opened with a mootness claim, which the Court challenged, but the government answered that while mootness could be an issue in election cases, it was not a problem in this case, where the plaintiff was responsible for many of the delays.

The government then noted the speculative future of the claim and the breadth of citizenship claims that might need to be raised. When the court later noted that its previous mootness case included the fact that Barack Obama was ineligible for another election, the government tried to raise the greater speculative nature of the political process. But the court pressed back that "capable of repetition yet evading review" inherently speculate about the future.

The court also mused that this exception often refers to the plaintiff's injury, as in the case with abortion cases; but, the court noted, if he has standing, what are the odds the plaintiff would face an election against someone born to foreign-born parents? That, the court said, might be a different case, as it seemed quite likely Mr. Paige would run again and face someone in that citizenship posture. But, another justice pressed back that it seemed like an advisory opinion, as it would be conditioned on "if" Mr. Rubio or someone else ran for office.

The court then asked about standing--a statute says that an election may be contested by any voter in the process. The government described the standing as something ministerial, such as the conduct of the election itself; or to matters like voter intimidation. But those are about the validity of the election, the government said, and the statute does not extend to challenges about qualifications.

The government moved back to the questions of Mr. Paige as a candidate challenging Messrs. Cruz and Rubio and argued that he was not truly a competitor in the election, such that he did not have standing.

The court started to worry about competitor standing--suppose Mr. Obama ran for governor in Vermont as a resident of Illinois? The government conceded that his general election opponent might have standing in that case.

All in all, the court muddled through a series of standing, mootness, and statutory issues, with virtually no time spent on the merits. It's unlikely we'll see any surprises... but only time will tell.

Slopegraph of electoral votes and popular votes for presidential candidates

After my perspective on electoral vote and popular vote margins--in which I argued that the popular vote is meaningless--I thought about how Electoral College and popular vote margins related to one another. I took a stab at a visualization by creating a slopegraph.

This was much more challenging than I thought. And perhaps it's more deceptive than informative. But why not give it a shot and let the critiques come....

I wanted to show the relationship between electoral votes and popular votes. I started by taking the raw popular vote totals of each candidate--this could have been as a percentage of electoral vote, but 1968 really screwed things up and messed with the visualization if I were using the raw electoral vote totals as the left data point, so I took the slightly less perfect version of the raw vote totals. I started from 1944, which had just 531 electoral votes, in comparison to today's 538, and some other deviations along the way.

Then I opted for the percentage of the two-party popular vote margin, which was also imperfect as a kind of comparison--it might lead to significant fluctuations if there is a particularly significant third-party candidate who draws votes disproportionately from one candidate.

View fullsize

In order to do the slopegraph on two different Y axes, I opted to calculate Z-scores for each side. That offered the relative performance between electoral votes and between popular votes, and it offered some comparable scale between the two from 1944 to 2016.

You can see a couple of significant differences between the electoral vote "landslides" of 1972 (Nixon winning 520 electoral votes, dark green) and 1980 (Reagan winning 489 electoral votes, light green). In '72, Nixon snagged a whopping 61.8% of the two-party popular vote. But in '80, Reagan secured just 55.3% of the two-party popular vote.

There's not much of a rhyme or reason between the performance in the Electoral College and the popular vote--except that we might notice particularly low-performing popular vote winners: Bush in 2000 (271 electoral votes, blue) had the razor-thin electoral advantage; somewhat healthier were Trump in 2016 (306 electoral votes, pending December 19, red) and Kennedy in 1960 (303 electoral votes, orange).

In any case, perhaps after all the flaws I've identified and the meaningless of the popular vote, anyway, such a slopegraph is of less than even marginal value. But here it is, if you find it of interest.

No, the Electoral College will not give the presidency to Hillary Clinton

There is a nascent but rapidly growing effort from supporters of Hillary Clinton to persuade presidential electors who would otherwise support Donald Trump to cast votes for Mrs. Clinton instead when the Electoral College meets December 19. Absent an extraordinary change of circumstances, it simply won't happen. Mr. Trump will win a majority of electoral votes on December 19 and become the 45th president of the United States.

It's worth noting that a lot of options to affect the presidential outcome have long since past--usually, waiting until after the election is not a good idea to affect an election.

I wrote back in March that state legislatures could choose their own electors instead of leaving the matter to a popular vote; but after a popular vote was held November 8, that strategy is not an option.

I also wrote in August that parties could select electors inclined to support their preferred candidate. The electors, however, have already been selected.

Instead, the only strategy for Mrs. Clinton's supporters is to turn to the Electoral College itself and persuade electors to be "faithless"--that is, persuade them to vote not for Mr. Trump, to whom they pledged (formally or informally) their support, but Mrs. Clinton.

First, it's worth noting that these are loyal Republicans who were selected as Trump electors. Many of them are loyal Trump supporters. The list of viable options, then, is limited to those who oppose Mr. Trump--and not just oppose him, but affirmatively prefer Mrs. Clinton (more on that point below). And this after Mr. Trump has won the election (at least, by all popular reports). It might be that Mr. Trump is not overly popular with many in the Republican establishment. But convincing them now to vote for someone else seems impossible.

Furthermore, these are electors in states that cast a plurality of their votes for Mr. Trump. Going to them and telling them to ignore the wishes of the voters in their own state for the wishes of the country as a whole--which, really, is overwhelmingly the wishes of California and New York--is even more unlikely.

Second, the electors would need to flip to Mrs. Clinton, and not simply refuse to vote for Mr. Trump. In order for a candidate to win, he must secure 270 electoral votes. If he fails to do so, the race is thrown to the House of Representatives, where each state receives one vote, and a majority of the states (26) is required to secure the presidency. Even if enough Trump electors threw all their votes to, say, Mitt Romney, no one would have a majority, the election would go to the House, and the Republican-controlled House where Republicans control a majority of state delegations would, in all likelihood, simply vote for Mr. Trump--absent yet another colossal effort to convince them to change their minds and somehow vote for Mrs. Clinton.

Third, the margin of victory is onerous for Mrs. Clinton's supporters. It appears Mr. Trump has won at least 290 electoral votes, meaning 21 electors would need to switch to Mrs. Clinton to deny him a majority, 22 electors to give her a majority, and 23 or 24 electors to account for Mrs. Clinton's own possible "faithless" electors. If he holds onto Michigan, she'll have secured 306 electoral votes, meaning the numbers increase to 37, 38, and 39 or 40.

These are Herculean numbers under almost any scenario. Consider that in the last 100 years, just nine (depending on your math) electors have been "faithless" and voted for someone other than the person pledged to support. Granted, no such concerted effort has been made to change electors' minds. Robert M. Alexander has surveyed presidential electors and discovered that serious lobbying efforts have occurred before, and that about 10% of electors in previous elections have considered voting for someone else--but did not do so.

Fourth, a few states purport to bind their electors to the individuals they are pledged to support. I've argued such laws may well be unconstitutional and should be repealed. But as they are on the books, it would either limit the pool of possible electors who could change their minds or stir litigation, possibly in multiple states, that would inspire even greater complexity, particularly if Congress is faced with multiple slates of electors.

In short, there is no realistic chance that the Electoral College will change the result of this election. This is different than saying it is not legally possible; as I've noted and defended repeatedly, electors are permitted to vote for whomever they desire--it is that there is essentially no likelihood enough of the would do so in such a way to change the outcome of the election. Circumstances change, of course, and something might still inspire a significant number of electors to change their minds and vote for someone else. But the odds are low. And we have fairly settled expectations that our electors will not be "faithless," something unlikely to change in the weeks ahead.

Could California vote for #Calexit? Probably not

Recently, "#Calexit" has been trending in California, a type of secessionist movement similar to Britain's exit ("Brexit") from the European Union. By popular vote, Britain approved the move, which has no legal effect according to the High Court but which continues to affect the political sphere.

Could Californians undertake a similar move and vote to "exit" the United States? The short answer is, probably not.

It's worth emphasizing I only examine whether California could vote to leave the United States--but there is some question on the merits I'll mention below. I defer to international law experts about the legality of such a move, but there is some history suggesting it could not, at least on its own, do so.

First, the #Calexit movement is not terribly sophisticated. It's worth noting this effort started months ago and has since been adopted as the new vessel for secessionists. It is not clear whether the proponents intend a ballot initiative, referendum, or an advisory question. (Following some of their comments on social media, proponents use terms like these interchangeably, if not randomly.) The form matters, which I'll broadly outline below.

A ballot initiative would not be permitted. Ballot initiatives in California include proposing new statutes or constitutional amendments. It is not clear that either could properly authorize secession from the United States. A constitutional convention might, I suppose; but that does not occur via initiative.

A referendum also would not be permitted, because it is, well, impossible. In California, a referendum is a decision by the people to ratify or reject a law enacted by the state legislature. Because there is no secessionist law that the state legislature has enacted, there is nothing for a referendum to do.

Instead, an advisory question would be the means to recommend #Calexit, and it would be that--a recommendation by the people, something like a public opinion poll but carries greater weight having come from the ballot box. Even that is limited.

For starters, a citizen-led advisory question is not permitted in California. In American Federation of Labor v. Eu (1984), the California Supreme Court held that the people had the initiative and referendum powers, but those powers extended only to those matters that enacted laws. Advisory questions were not authorized. It explained:

We acknowledge the arguments of the proponents that there may be value to permitting the people by direct vote not only to adopt statutes, but also to adopt resolutions, declare policy, and make known their views upon matters of statewide, national, or even international concern. Such initiatives, while not having the force of law, could nevertheless guide the lawmakers in future decisions. Indeed it may well be that the declaration of broad statements of policy is a more suitable use for the initiative than the enactment of detailed and technical statutes. Under the terms of the California Constitution, however, the initiative does not serve those hortatory objectives; it functions instead as a reserved legislative power, a method of enacting statutory law. The present initiative does not conform to that model.

But a precedent has now been set in California on a different type of advisory question. The California legislature passed what would ultimately become Proposition 59 on the 2016 general election ballot. Some litigation kept a similar proposal off the 2014 ballot, but after full briefing the California Supreme Court considered whether the legislature could refer an advisory question to the people via initiative--here, a call to California's elected officials to lead an effort to amend the United States Constitution to permit greater regulation of campaign finance.

In Howard Jarvis Taxpayers Ass'n v. Padilla, the California Supreme Court permitted such an advisory question to appear on the ballot. The Court accepted the argument that the legislature had "the inherent power to conduct an investigation in order to select the wisest policy course." It could then refer such questions to the people as a part of its investigatory power. But the Court was careful to limit this power: among other things, "the investigative power permits inquiry only into those subjects 'in reference to which [the Legislature] has power to act.'"

An advisory question, then, could not ask for California to secede if the legislature lacked the power to secede. Instead, it could only ask, along the lines of Proposition 59, to urge elected officials to pursue (amicable?) secession with the federal government. That is, unless (and this is on the merits, as mentioned above) the state legislature does have some right to do so, but that is a much more complicated question--even though, I think, the answer is probably no.

That said, the (dare I call them unsophisticated) claims from the current #Calexit movement suggest they will be gathering signatures for this ballot measure, which suggests they do not intend to have the legislature refer the advisory question to the people. (It's also deeply unlikely that the state legislature, if asked, would do so.) Proponents apparently intend to simply gather signatures. And whether they do so as a ballot initiative (which the people lack to enact as a matter of law) or an advisory question (which the people lack the authority to do under the initiative power), the effort would likely fail, and the people could not vote for #Calexit.

UPDATE: The "Yes! California" movement helpfully directed me to their proposal they filed last year. It is an initiative that requires a regular referendum to call for secession, and in the event of a referendum that approves secession, triggers obligations from the California government to pursue good-faith secession negotiations with the federal government. I am fairly confident that laws that purport to bind future legislatures (if not all government officials) to act in a particular way run afoul of basic principles of legislative autonomy. (See generally People's Advocate v. Superior Court (Cal. App. 1986).) Finally, such a transformative change in California is likely a "revision" to the state Constitution, not simply an "amendment," and as such would need to come from the legislature and not an initiative petition.

How recounts in presidential elections work under Pennsylvania, New Hampshire law

Tonight, or this morning, a number of elections are particularly close. From my own guess, I think the margins in Pennsylvania (20 electoral votes) and New Hampshire (4 electoral votes) look to be the closest. In the event that these elections prove to be decisive in a presidential candidate's quest to secure 270 electoral votes, recounts are likely. (Of course, if a candidate concedes, or if the margins become unrealistically large, or if there are still more tipping point states, or there are others, this prediction is sure to be wrong!)

Many recount procedures begin administratively through the Secretary of State or an election board. It's worth noting with any contested election that usual basis for filing a claim is in state court. For all the attention given to federal causes of action, the ordinary act of recounting is typically reserved to state courts. In Florida in 2000, the federal issues were raised as a counter to the procedures implemented by state courts; they were not originally federal causes of action. It's likely, then, that we would expect basic recount procedures to arise in state court; federal claims could arise in that context. It's less likely that claims would originally be filed in federal court, but there are plausible scenarios where it might happen (and where federal courts would not abstain from exercising jurisdiction, another issue!), but I'll focus on the state-based claims at the moment.

It's also worth noting that after Bush v. Gore, the key deadline this year is December 13. That's the date set by federal law for presuming the regularity of the election results in Congress. Most opinions in Bush v. Gore recognized that this deadline was an important date for a state to meet (although some disputed how important or what procedures should take place in the late days before that deadline). Any litigation, then, would occur with an eye toward that date.

Below are the general timing and triggers for basic recounts, but there are obviously many more conditions and opportunities that can arise beyond these. But as we hardly know if there will be challenges, much less here, I'll stick with a couple of bare outlines of order and timing.

Pennsylvania

An automatic recount is triggered under Pennsylvania law if the margin of victory is one-half a percentage point or less. (25 Pa. Stat. § 3154(g)(1)(ii)) (There is no recount if the losing candidate requests no recount.) The Secretary issues an order by 5 pm ET November 17. (25 Pa. Stat. § 3254(g)(2)) The recount and recanvass must be finished by noon ET November 29. (25 Pa. Stat. § 3254(g)(5)) (Some more background is here.)

New Hampshire

There are no "automatic" recounts in New Hampshire, but parties can petition for a recount if the margin is less than 20%. (N.H. Rev. Stat. § 660:1) (Some more background is here.) A candidate must file a petition by November 11. The process to recount must begin by November 16. (N.H. Rev. Stat. § 660:4)