Towards an Unelected President

The exciting conclusion of Some Constitutional Amendments #5.

Many writers propose constitutional amendments in order to demonstrate their fantasy vision of the perfect regime. In Some Constitutional Amendments, I propose realistic amendments to the Constitution aimed at improving the structure of the U.S. national government, without addressing substantive issues.

This is Part II of the discussion of the Electoral College. In Part I, we saw how the Founders’ original plan for presidential nomination by the Electoral College and election by the House of Representatives fell apart, what we improvised in its place, and observed how bad it is.

---

“Strange women lying in ponds distributing swords is no basis for a system of government! Supreme executive power derives from a mandate from the masses, not from some farcical aquatic ceremony! I mean, if I went around saying I was emperor just because some moistened bint’d lobbed a scimitar at me, they'd put me away!”

President Lincoln said, “A majority held in restraint by constitutional checks and limitations, and always changing easily with deliberate changes of popular opinions and sentiments, is the only true sovereign of a free people.” As we saw in Part I, there is no shortage of popular opinions and sentiments in our presidential election system, but we have lost every constitutional check and limitation save one: the flapping ghost of the original Electoral College—and all that does today is sometimes randomly award the presidency to the person with fewer votes. The limitations, the moderating institutions, are all gone.

At the end of Part I, I invited you to think about how you might repair our broken presidential election system. How might we increase the odds that the most powerful man in the world will be a human worthy of the office? How can we ensure that he is loyal to the rule of law, competent, and a man of reasonably sound public character? These are hard questions.

Even harder: two centuries of practice have taught us that the answer to these questions is always to wrest more power away from “elite” institutions and hand it back to the People. As we saw last time, this totalizing democratic instinct is what drove us into the ditch we’re in right now! The democratic will of We, the People is essential to our freedom and flourishing, but, unless that nuclear furnace of popular opinion is moderated by institutions, you end up with polarization, misinformation, and billion-dollar campaigns decided by negative advertising and the Chinese Communist Party.¹ The Parties, not the People, take power. We have to stop thinking like populists and start thinking like Founders.

You may have done a lot of hard thinking these past couple of weeks about presidential elections. I didn’t. I think the Founders basically got this right the first time. I’ll repeat something I referenced in Part I—Hamilton in Federalist 68—because it cuts so neatly to the heart of the matter:

It was desireable, that the sense of the people should operate in the choice of the person to whom so important a trust was to be confided. This end will be answered by committing the right of making it, not to any pre-established body, but to men, chosen by the people…

It was equally desirable, that the immediate election should be made by men most capable of analizing the qualities adapted to the station, and acting under circumstances favourable to deliberation and to a judicious combination of all the reasons and inducements, which were proper to govern their choice. A small number of persons, selected by their fellow citizens from the general mass, will be most likely to possess the information and discernment requisite to so complicated an investigation.

It was also peculiarly desirable, to afford as little opportunity as possible to tumult and disorder. This evil was not least to be dreaded in the election of a magistrate, who was to have so important an agency in the administration of the government, as the president of the United States… this detached and divided situation will expose [the electors] much less to heats and ferments, which might be communicated from them to the people…

Above all else, we must avoid a mass national election, which is why the Founders tried to short-circuit ours. Instead, the President should be chosen by a deliberative body of only a few people, who should be selected for their practical political wisdom. They are best suited to nominate good candidates for President. The Founders recognized this, and history bore them out: the Congressional Nominating Caucus employed from 1800-1824 did an extraordinary job nominating presidential candidates of the highest caliber. The caucus-convention system and its “smoke-filled rooms,” which filled the nominating role for the next century and a half, did a pretty good job, too.

By contrast, the presidential primary system put voters directly in the driver’s seat. Primaries started to overtake the caucus-convention system in 1972 and had effectively killed the old system by 2013. The first thing the new system did after triumphing was give us Donald Trump vs. Hillary Clinton, the number-one and number-two least popular presidential candidates in American polling history. I don’t think that’s a coincidence. When we put in place a system calibrated to reward populist, polarizing demagogues who cared only about governing the half of the country that voted for them, we lost the right to be surprised when we got populist, polarizing demagogues who cared only about governing the half of the country that voted for them.² The seeds were already planted and visibly bearing fruit by the time of the Obama-Romney face-off. 2016 simply showed us their full flower.

We, the People, are very good at a very great many things, which is why America has been such a success. We are not good at winnowing the 330 million potential presidents in this country down to just 2 candidates by ourselves. The process of choosing between the 2 is no picnic, either. (“Heats and ferments” indeed!)

Heck with all that. I still can’t say it better than Hamilton (one! more! time!):

The immediate election should be made by men most capable of analizing the qualities adapted to the station, and acting under circumstances favourable to deliberation and to a judicious combination of all the reasons and inducements.

Constitutional politics is a matter of prudence, not a matter of proof. I cannot give you a syllogism showing that Hamilton was right, and it may cut against every democratic instinct you’ve developed. How can the most powerful officer in our whole democratic system not be directly accountable to voters?

I answer: “But John Roberts and Janet Yellen have never been accountable to voters!”

Okay, I joke, but I convey an important truth: we already have very powerful offices that aren’t directly elected, and we understand why that independence is good and necessary. All I’m saying here is that the Presidency is one of those offices.

Imagine a competent President of reasonably good character who didn’t have to worry about the polls, who didn’t have to Deliver on Campaign Promises, and who (flip side of the coin) enjoyed no democratic “mandate” to try to force new domestic policy through Congress. The unelected President would have little choice but to faithfully execute the laws of Congress and deal with foreign policy, which is all the Presidency was ever intended to do anyway. Congress, meanwhile, would lose its current (strong) incentives to prevent an opposite-party president from being successful. Congressmen could simply… govern as they and their constituents see fit? Crazy talk, I know.

Of course, just because the President is not directly elected does not mean the voters get no say at all. On the contrary, in Hamilton’s view (and mine), the American People need to have a strong voice in who these presidential electors are. The People’s horse-sense is too useful a check and balance against elite obsessions to exclude from the process, and their sovereignty too directly tied to the President for them to be denied any say. However, once the People have chosen the electors, the electors need to lock themselves up in a room and hash it out face-to-face, through careful discussion, and vote their consciences. (Filling the room with smoke is strictly optional.) They must not be under our direct control, or we will succumb (again and again) to electing demagogues and tyrants to the White House.

Waiting for Supermen

However, as you recall, the Founders tried to establish this exact system, deliberation with popular input—and, as we saw in Part I, they failed! The Electoral College never worked!

Because of Duverger’s Law, two and only two political parties naturally came into being soon after the Founding. Protection of party interests meant the electors could not vote their consciences. Instead, they had no choice but to coordinate their votes. Any honest vote was so useless it might as well have been thrown away, so coordination became essential. Someone had to fulfill this coordinating role, but, because the presidential electors only met once (in thirteen different state capitols, at that), it couldn’t be the electors themselves! Inevitably, the parties improvised a solution, and the Congressional nominating caucuses were born.

Once nomination was taken out of the Electoral College’s hands, the electors played no deliberative role anymore. They quickly came to be elected or appointed based on party loyalty alone, as mere functionaries. Hamilton’s “most capable” men? Hah! Presidential electoral slots are doled out as a reward for party service, no more, no less. (Trust me; I’ve voted for presidential elector candidates in district party nominating conventions.) Soon enough, electors didn’t even have the legitimacy to exercise true deliberation. When electors dared to exercise the discretion with which they were originally entrusted, we derided them as “faithless” electors!

What we should learn from the failure of the Electoral College is that, while it is highly desirable to have a deliberative, representative body, elected by the people, to nominate and elect the presidential candidates, there are certain things that body can’t be:

• It can’t be a dislocated body that meets in a bunch of different places at once and casts separate votes; the body has to meet in order to actually deliberate, or it will be overpowered by the Parties’ nominee-coordination power on one side and the House’s contingent-election power on the other.

• It can’t vote just once, but has to be given a chance to deliberate, vote, deliberate, and vote some more—like a jury. Otherwise, tactically coordinating one’s single electoral vote will become too important, and the parties will fulfill that coordinating function.

• Nor can the members of the body be beholden to political parties, such that parties can effectively punish members for voting the “wrong” way (and thus impose their own nominations). The members of the revised Electoral College must be able to vote freely, which means they must be able to stand up to the parties, which means the role of presidential elector can’t be a sinecure you give to Janet, the sweet lady who’s distributed party pamphlets door-to-door since the Nixon re-election campaign.

• Ideally, the members of the body are not members of the federal legislative or judicial branches, because then the President would owe his office to other branches, which could make him subordinate to them. The Founders were very worried about this dynamic, and, as we saw from the persistent criticism of the Congressional Nominating Caucuses, so were early Americans.

• Finally, as the Twelfth Amendment debacle taught us, we can’t put president and vice-president on the same ballot in a way that forces risky tactical voting! That was just embarrassing!

So where do I propose to find these superhuman presidential electors, who manage to represent the People (including the parties) without being owned by the parties? What group of electors could be large enough to represent the diversity of our vast nation while being small enough—and sane enough—to be truly deliberative? What group of electors, outside the rival federal branches, truly understands what the nation needs from a chief executive?

I think I’ve beat around the bush long enough.

It’s Governors!

The Founding Fathers never considered governors in their electoral scheme, because governors circa 1789 were weak-kneed dinghies of suckitude.³ Governors were typically elected for chaotic, single-year terms. They were often term-limited to just one or maybe three total years. They were tightly controlled by their state legislatures through executive councils and other devices. Most governors were elected directly by their legislatures. Of thirteen states, eleven had no gubernatorial veto power at all.⁴ Early Americans’ (understandably) feared that any strong executive would eventually turn himself into a king, so they guarded executive power jealously. There was even a popular idiom, “Where annual elections end, tyranny begins.” As a result, though, post-Revolution governors were nothing to write home about. Pennsylvania distrusted executive power so much, they didn’t even have a governor. They had a “Supreme Executive Council.”

Governors today are a rather different story. A governor today serves four years,⁵ and they all have a broad veto power. Some wield the mighty line-item veto, a tool of executive power so great that surely even the Federalist Founders would have spurned it. Governors are, nearly always, the undisputed masters of their state parties; their vast power to dole out policy and employment favors gives them massive leverage against the party.⁶ They are, very frequently, the most respected and admired political officers in the American firmament. Twelve governors currently have approval ratings over 60%, an approval rating no president has held since the Obama honeymoon in 2009. The least popular governor in America, Tina Kotek of Oregon, still has a positive net approval rating, which is more than you can say for pretty much any national political figure.

Governors have to win popular elections in vast states, so, like anyone who has to persuade more than half a million voters, governors are generally demagogues, not saints. However, governors are rarely kooks. Kooks can win election to Congress, even the Senate. They can be appointed to courts for life. State legislatures are positively riddled with kooks. This is largely to the good; kooks are part of our national life and it is good that they, too, are represented in our legislatures. But it’s pretty tricky to get an entire state to elect you governor when you’re a kook. It does happen! I’m from Minnesota, home of Gov. Jesse “the Body” Ventura. Kari Lake is a kook, and came within a hair’s breadth of being governor of Arizona—but, even in that very red state, she couldn’t close the deal, ‘cause she’s a kook. For the most part, voters are too worried about what a bad governor will do once in power to elect a real kook.

I’m not saying governors are terrific role models of ideal citizenship. But I think there’s a strong case to be made that, on average, governors are the least bad group of senior elected officials in the United States.

Governors are certainly the elected officials with the most executive experience, which means they have the best understanding of the challenges the President will face and the qualities that the President will need to succeed. Moreover, unlike some people in Congress, governors can rarely build a career on stubbornly opposing everything. Governors tend to become skilled negotiators and diplomats.

What did Hamilton say? “[M]en most capable of analizing the qualities adapted to the station, and acting under circumstances favourable to deliberation”? Sounds like governors fit the bill! Let’s let them pick the President.

Every four Decembers, have all fifty state governors meet somewhere. They argue. They vote. If there’s no President, they vote again. They keep on voting until there’s a President. Badda-bing badda-bang.

(The spirit of Amendment XXIII says we should count the mayor of the District Columbia as a governor. Fine. Fifty-one governors, then.⁷)

The Double-Majority Rule

Of course, it’s not that simple. Some states are small and other states are large. If we gave every state governor an equal vote in the Gubernatorial Electoral College, the smallest 26 states could elect a president over the 24 largest states… even though the smallest states contain only 16% of the U.S. population! Wyoming has barely 1% the population of California, but this would give Wyoming just as much say in the presidency as California!

This is exactly why Democrats are terrified of Article II’s unit rule. As we discussed in Part I, the unit rule gives every state in the House of Representatives a single, equally-weighted vote whenever the House has to vote directly for President. The unit rule favors small states (which is why small states insisted on it at the Constitutional Convention), and small states in 2024 tend to vote Republican.

We don’t want the People voting directly for the president, but we do want the will of the People broadly represented in the presidential election. If small states invariably control the outcome of every presidential election, our system won’t be representative. Also, as a practical matter, Hell will freeze over before large states agree to it, just as small states will never allow a direct popular election for President. The governors’ votes in the Gubernatorial Electoral College need to be weighted by population, so that the American population is effectively represented.

At the same time, the states are important, too. The President is chief magistrate, not of the United People of America, but of the United States of America. His government was instituted by states, and it serves the People through the states. Large states and small states, in particular, have different cultures and different needs, both of which need representation. If we simply weighted each governor’s votes by population, the vote of California’s governor would be worth sixty-nine times more (nice) than the vote of Wyoming’s governor. In fact, the nine most populous states could decide the presidency themselves, without input from the other forty-one! That ain’t quite the right balance, either. Also, as a practical matter, Hell will freeze over before the other forty-one states allow it.

For this reason, I propose a double-majority rule: in the Gubernatorial Electoral College, a candidate shall be elected President if the governors voting for her represent both an absolute majority of states (26) and an absolute majority of the U.S. population (167 million people). This ensures that the President is a consensus candidate who enjoys broad support. Small states cannot dictate to large states, nor large states to small states.⁸

It is worth pausing for just a moment to consider what an election under this rule might look like. Right now, today, Republicans have 27 governorships (53%), while Democrats have 24 (47%)⁹. On the other hand, Democrats represent 183 million people (55%), while Republican governors represent only 151 million voters (45%). If an electoral convention were held today, the governors would not be able to elect a President without winning at least a few votes from the other party. This turns moderates into kingmakers: Democrats, who need two more states to elect, will have to find a presidential candidate, somewhere in America, who appeals to people like Chris Sununu (R-NH) and Phil Scott (R-VT). Meanwhile, Republicans, who are 15 million votes short, will have to find someone who can win support from the likes of Laura Kelly (D-KS), Andy Beshear (D-KY), and Roy Cooper (D-NC). It is inconceivable that an election under these constraints would put the likes of Trump or Biden back on the throne.

Of course, there are going to be cycles where one party achieves a double majority by itself. Republicans last achieved this in 2016 (when they had 31 governorships and represented 61% of Americans). Democrats last did it in 2008 (when they had 29 governorships and represented 54% of Americans). You might think, “Ah, but then it all falls apart! In those cycles, purely partisan picks—or at least some nutcase like Bernie Sanders or Ted Cruz—will get elected anyway!” It’s true that, as America shifts redder or bluer, we are likely to get redder or bluer presidents. That’s a good thing! However, even when that happens, it seems unlikely that the Gubernatorial Electoral College will elect a kook. As we’ve seen, governors are less beholden to the political parties than just about anyone else in our system. Take 2016, when Republicans had a double majority: do you really think John Kasich (R-OH) and Bruce Rauner (R-IL) would ever, in a million years, vote to make Ted Cruz the President of the United States? In the real 2016 election, Kasich was so desperate to stop Cruz that he accidentally drove the GOP into the hands of Trump instead. If the Gubernatorial Electoral College had been online, we likely would have ended up with some popular, mainstream Republican like Scott Walker (R-WI) in office, not Donald Trump.

So, yes, under the Gubernatorial Electoral College, your party will sometimes still lose decisively, and that will hurt (Democrats loathe Walker)… but, even then, the breadth of consensus necessary to carry the College means it will hurt a lot less (Walker’s a lot less bad than Trump!).

Deadlock, Shmeadlock

There is, of course, the risk of deadlock. Sometimes, an electoral body is deeply divided and just can’t find its way to agree on a single, broadly-acceptable candidate. Under our current system, voters can’t search for a compromise candidate and are forced to pick one of the two major-party candidates (lest we throw our votes away). Whoever gets a mere plurality wins (usually, more or less). We end up with unacceptable choices being elected with less than majority support.¹⁰

A deliberative electoral body doesn’t have to do that, and actually can’t. They have to keep looking until they find a broadly acceptable candidate. However, that can take a while. Remember the 2023 Speaker of the House election? That took eight days and eighteen ballots to resolve.¹¹ Even under our current system, a tie vote in the Electoral College throws the election to the House of Representatives, which has just two weeks to elect a new President before the current President’s term expires. Deadlock was a real danger until the ratification of the Twentieth Amendment, which created rules for presidential succession during deadlock.¹² When the electors also have a super-majority or double-majority rule, requiring even broader consensus, that can make deadlock harder to break.

When I think “super-majority election,” two organizations immediately pop into my head: the Democratic National Convention of 1832-1932, which required a two-thirds majority of delegates¹³ to nominate their presidential candidate; and the Roman Catholic Church, which, from 1179¹⁴ until 1996,¹⁵ required a two-thirds majority for the conclave to elect a new Pope. And, you know what? Some of these elections got pretty hairy!

In 1924, the Democratic Establishment’s preferred candidate, Al Smith, deadlocked against the Ku Klux Klan’s preferred candidate, William G. McAdoo. Although neither controlled a majority of the delegates, they both controlled more than one-third—which, under the rules, was enough for them to veto one another. The convention, forever known as “the Klanbake,” took one hundred and three ballots, over sixteen sweltering summer days in Madison Square Garden, to finally nominate compromise candidate John Davis.¹⁶

Still, this was nothing compared to the deadlock in the papal conclave of 1292. The ten voting cardinals were split between supporters of the Orsini family and the Colonna family. After ten days of balloting in April, they shrugged and adjourned for three months, popeless. Then, an epidemic caused them to adjourn again… this time for a year! They gathered in Perugia in fall 1293, but they seem to have met infrequently, carried on their ordinary episcopal business while in Perugia, and remained deadlocked. King Charles II stopped in town in March 1294 to yell at them, but Cardinal Caetani (future Pope Boniface VIII; not a fan) bleated back that the election was to be free of royal coercion, thank you very much, so the election dragged on all the way into summer 1294. They were about to adjourn again when an impassioned letter from local hermit Pietro de Murronne urged them to pick someone… so the cardinals picked Pietro, who reluctantly became Pope St. Celestine V. The papal deadlock had lasted more than two years.

The Klanbake was the longest political convention in American history. It was a slog, but it got the job done: the stubborn delegates eventually recognized that a compromise candidate was needed, and promptly elected him. Unable to adjourn or resume their ordinary lives until they had a nominee, they figured it out in just over two weeks. The election of Celestine V reached the same conclusion—they needed to elect a compromise candidate—but it took them two years to accept this. It seems they dallied so long at least partly because they were able to go about their regular business during the election. They faced no threat of discomfort if the election dragged on, nor any possibility of anyone assuming the office without their say-so. For just this reason, every papal election since 1294 has taken the form of a conclave, where the cardinal-electors are held under lock and key until “Habemus papam!” Both cases are extreme outliers, but the contrast between them teaches us a useful lesson: if you want to be sure the electors get the job done, don’t let them leave.

The Twentieth Amendment ensures an orderly continuance of American government no matter what happens (which creates its own pressure on the governors). However, to encourage the governors to identify a compromise candidate sooner instead of later, we should ensure they remain secluded until they reach a decision. The Gubernatorial Electoral College should generally be able to set its own internal rules and procedures, but, to ensure that they stay on task, we will allow Congress to set the regulations for their sequestration.

There is one more problem to deal with.

Republican Insulation

Before the Seventeenth Amendment, U.S. Senators were elected by state legislatures, not popular mass elections. This was supposed to guarantee (among other things) that Senators would be cool-headed deliberators who impressed other legislators, not populist demagogues who knew how to rile a crowd. (The Founders recognized that an element of populism was essential to representative government, but populism was to be seated in the House, not the Senate.)

However, I described in Part I how the small-d democratic impulse is like a river running past a mountain, slowly wearing away every small-r republican institution that might stand in its way. During the nineteenth century, the original plan for electing senators broke down. A weird, degenerate system for electing U.S. Senators emerged in its stead.

No doubt you’ve heard of the Lincoln-Douglas debates, where Abraham Lincoln and Stephen Douglas, both running to be the next U.S. Senator from Illinois, clashed over slavery, sovereignty, and the Supreme Court in front of raucous throngs. Famously, Lincoln won the debates but lost the election—until his rematch with Douglas just two years later, when they ran against one another for President of the United States.

But, hang on, go back a sec. They were running for U.S. Senate. The Seventeenth Amendment was decades away; the next senator would be chosen by the Illinois legislature. There was no general election for Lincoln to lose! So what were Lincoln and Douglas doing out there? Why were they making their cases to the teeming masses, when the people who would actually pick the senator were all in Springfield? What, to coin a phrase, the heck?

In 1858 Illinois, as in many other places by 1858, Senate candidates were chosen by the parties during summer of the election year. (Lincoln was nominated by the Republican State Convention on June 16.) The two candidates then became the “face” of their party for the next few months, leading up to the fall elections for state legislature. All the Republican candidates for legislature promised to vote for Lincoln for Senate. All the Democrats promised they’d vote for Douglas instead. The Senate election was the leading issue in the campaign (as it often was). When the People went to the polls on January 5,¹⁷ they were formally voting for their new state legislators, but, in actuality, many of them were simply trying to help their favorite U.S Senate candidate win the U.S. Senate election. The legislative election had become a proxy election for Senate.

This happens, sometimes, when popular interest in an indirectly elected office becomes too strong. For example, some people vote for President as a proxy election for the next Supreme Court justice. (I’m often one of them!) Indeed, our starting point for this whole discussion is that nobody today actually votes for President at all. We are all formally voting for “men most capable” to hold the office of presidential elector. However, that became a proxy election so long ago that, in most states, the electors’ names no longer even appear on the ballot.

Proxy elections do damage twice over. First, they take decisions that the Constitution confided to some deliberative republican body—usually for good reason—and give those decisions over to the mob… or, perhaps more accurately, to the Parties. In 1850s Illinois, their Senators, who were supposed to be insulated from popular pressure as a check against the populist House, ended up owing their seats to that very same populism.¹⁸ Second, proxy elections distort the proxied elections themselves. Your state legislative election should be the quintessential local-issues campaign, where you pick the guy best able to fight for your town’s interests in the Capitol. Instead, the election ends up being about slavery hundreds of miles away, and you pick the guy you deem most loyal to Honest Abe.

Although our governors are (as I have argued) the “least bad” class of government officials now, it’s all too easy to imagine that, if governors gained the power to elect the President, immense popular interest in the presidential election could turn gubernatorial elections into pure proxy elections (c.f. Campbell’s Law). The party primaries might go on uninterrupted, yielding up the same terrible nominees; voters in gubernatorial elections would stop caring about what their states need and elect their governors based on national issues; governors would be nominated for their loyalty to the party candidate, and perhaps even legally bound by state law to cast their vote for that candidate; we would end up with similar presidents and worse governors. This would be a bad outcome!

Fortunately, our proposed rule of double majority, with the inevitability of cross-party compromise, largely prevents this from happening. Even if governors were bound to vote for a certain party on the first ballot at the electoral convention, presidents will rarely be elected on the first ballot. On subsequent ballots, governors would be free to deliberate and seek compromise as usual. In any election where the parties are forced to compromise, a primary campaign would not only be useless; an open endorsement would likely be actively harmful to the nominees’ chances of becoming the winning compromise candidate. Nevertheless, we can strengthen the governors’ hands by adding a provision to the Constitution that should have been added centuries ago: an express prohibition on “binding” presidential electors in any way.

Moreover, to free the governors to debate freely and honestly, without fear of giving a too-honest soundbite to C-SPAN, we should give them authority to keep their proceedings private and their ballots secret. Zealous transparency has done a lot of damage to our legislative branch’s deliberative capacity in recent years. Many of the good things Congress has accomplished recently were thanks to the rise of Secret Congress, which occurs out of the limelight. Congressional transparency exists to make politicians accountable to special interest groups, political party apparatchiks, and (last and least) voters. We don’t want the Gubernatorial Electoral College to be accountable to those groups after the fact. The People should influence the presidential election process by putting people in office who are good at governing. That’s all. Any further popular pressure would short-circuit honest deliberation, which is the very thing we are trying to foster! We should never have C-SPAN cameras in this convention (much as I would enjoy watching it). If the convention finds it good or useful to publish its proceedings, so be it, but that decision should be left to the convention.

Even with these fixes, if there remains even a small possibility that your vote for governor today will decide who becomes President next month, there remains a risk that, in the heat of the moment, state voters will vote for governor based on national politics. (That is, even more than they already do!)

We can greatly reduce this tendency by creating a firm separation in time between the gubernatorial and presidential elections. Swing voter Joe Sixpack might very well vote for a governor based on his presidential preference if the governors are convening very soon (say, in three weeks), especially if the media is already abuzz with speculation about the leading presidential papabile. Joe is much less likely to let the presidency dictate his vote for governor if the electoral convention doesn’t start for a long time—let’s say eighteen months. The presidential race is too fluid, too frequently turned on its head by events, for swing voters to lock themselves into a candidate, or even a party, that far in advance. If there were a delay between gubernatorial elections and presidential ones, governors would be able to keep their elections about state issues, and they would be able to go into the presidential electoral convention free of any kind of democratic mandate to vote in any particular way. We will require, then, that governors who serve in the electoral convention cannot have been elected in the eighteen months leading up to the convention.

Voters (or partisan state legislators) may still ignore the deliberative function of the electoral convention. They could try to impose some kind of “advisory” instructions on their governor, demanding that he vote for a specific candidate (probably a partisan nominee) under threat of removing the governor from office if he fails to follow those instructions. We should prohibit this, and, for good measure, we should bar unscheduled elections for six months after, in case the legislature tries to make its displeasure known by triggering a recall election.

Technicalities

There’s just a few technical issues remaining.

We want the governors at the electoral convention, and we will naturally empower Congress to compel their presence if they try to shirk it. However, what if the governor dies, becomes disabled, or gets impeached and removed before the convention starts? We will need to allow the governor to choose alternates. They, too, will need to be chosen many months in advance, just like the governor, to ensure that their selection isn’t influenced by an ongoing presidential campaign.

What if a state has no governor? I think it’s been centuries since this happened, but Pennsylvania used to be run by a twelve-person Supreme Executive Council, and the Constitution therefore always refers to “the executive authority” of a given state, not to its “governor.” I think this is easily solved, as long as we notice the problem: we’ll just let each state’s “executive authority” designate one of its members to be that state’s elector.

It’s always been possible for the President-Elect to refuse election, but the public nature of modern campaigns has made it unlikely in the extreme. The Twentieth Amendment doesn’t really contemplate the possibility (unless you squint). The conclave-like nature of the Gubernatorial Electoral College means the governors might actually elect someone who wasn’t expecting it—possibly even someone who would refuse. We should account for this by giving the governors another bite at the apple, instead of falling back entirely on the Twentieth Amendment.

One oddity of the proposed amendment, specifically its rule that governor-electors must have been elected at least eighteen months before the presidential convention, is that it gives states strong incentives to move their gubernatorial elections to the first or second year after a presidential election. If they don’t, it is possible—even likely—that the governor they send to the electoral convention will sometimes be a lame duck, or possibly even an ex-governor. This is odd, but “odd” is well worth it when it buys us so much insulation against populist pressures. The much harder problem I had drafting this was figuring out how to avoid wrecking the two-year terms for governors in Vermont and New Hampshire, but I’m pleased to report that, in the end, I got the dates to line up, so Vermont and New Hampshire will not have to move their elections or amend their constitutions if this proposal passes.

The amendment does nothing to change the constitutional qualifications for office: the President must be at least thirty-five years old, a natural-born citizen of the United States, not an oathbreaking insurrectionist, and can’t have already served two terms. The governors here become the principal judges of these qualifications, although a disqualified President could be turfed out on a Congressionally-authorized writ of quo warranto. None of this is significantly different from how it works under the current system (if anything, it’s slightly clearer), but presidential qualifications and the mechanics of disqualification have been on my mind a lot lately, and perhaps on yours, too, so I figured I’d at least mention them.

I already said that our amendment should treat D.C.’s mayor as a governor. It is possible, however, that Congress will eventually turn most of D.C. into a state. (I don’t think they should, but that’s a separate blog post. The point is, they might.) Our amendment should not “break” if D.C. becomes a state, so we’ll need to account for that. D.C. might also be mostly ceded back to Virginia and Maryland (this is my preference). We should account for that, too.

There’s a similar problem with the already-existing Twenty-Third Amendment, which gives D.C. three electoral votes. If D.C. ever becomes a state, it will leave behind a rump federal district whose permanent population is, basically, the President and his family, and it is not clear what happens then to D.C.’s three electoral votes. This creates thorny legal issues for D.C. statehood. I oppose D.C. statehood, but this is silly and we should fix it while we’re in the neighborhood. (I suppose someone who’s really into D.C. statehood might support this whole amendment just for that!)

One final feature worth noting: our proposal eliminates contingent elections in the House, with their attendant unit-rule weirdness. This is good news for Democrats, who are terrified of a tie vote in the current Electoral College leading to a unit-rule election where small states are disproportionately powerful. Even Democrats who really don’t like the general direction of this whole proposal will, I hope, agree that it improves on the current system in several ways that matter a lot to Democrats—and this proposal, unlike a national popular vote, could actually get ratified someday. (Republicans should support this proposal because they already support the electoral college, and this proposal improves on it in every way.)

Finally, Text!

Enough! I’ve taken plenty of your time! Let’s spell it out:

Amendment XXXII

Section 1.

The President and Vice-President shall be elected as follows:

In each December prior to a year in which a presidential term shall begin, the designated elector of each of the several states shall convene at a time and place fixed by Congress, sequestered under such rules as Congress may, by law, prescribe.

During June of the year preceding the convention, the executive authority of each state shall prepare a list containing the designated elector and alternates. The elector shall be the current supreme executive authority of the state (or, if the state does not vest supreme executive authority in a single person, then a holder of the office or member of the body in which the state’s supreme executive authority rests). The list may contain up to four alternates, who shall be chosen, at the sole discretion of the state’s executive authority, but with the consent of each alternate, to serve as elector in the event of the prior elector’s death, disability, disqualification, conviction for crime, or removal by impeachment from all public offices. On that list, the executive authority shall choose and clearly identify the priority in which alternates should be called upon at need.

The executive authority shall then sign, certify, and transmit the list to the seat of government of the United States, directed to the Speaker of the House, who shall enter the names and their priority upon the journal, to be called upon at the need of the convention. The House of Representatives is authorized to compel the attendance of electors at the convention, in such manner, and under such penalties, as it may provide.

Section 2.

From among its members, the electoral convention shall elect a presiding officer and tellers, and shall determine the rules of its own proceedings, including whether and in what form its proceedings shall be published.

The acts of the convention are valid only if agreed to by electors representing a majority of the several states; and only if the total population of those states, as determined by the most recent decennial census, is at least half the total population of all states represented at the convention.

The convention shall elect a President of the United States and a Vice-President of the United States, then dissolve. Should the President-elect refuse the office prior to the start of his term, Congress may, by law, provide for the electors to convene again.

Section 3.

Any pledge, vow, oath, or any other commitment by an elector regarding his decisions in the electoral convention shall be null, void, and utterly without force from the moment it is made (excepting her oath to this Constitution). Any instruction, advice, or requirement laid upon an elector, outside the provisions of this Constitution, shall be likewise null and void.

No state may penalize, reward, or question any elector for his decisions at the electoral convention. No elector who holds any public trust may be subject to unscheduled election for six months after the start of the convention.

Section 4.

For the purposes of this article, the District constituting the seat of Government of the United States is considered a state.

Whenever the total area of said District is less than one thousand hectares, both this section and the entire twenty-third article of amendment to this Constitution are suspended.

Deals Worth Taking

I am optimistic that, someday, there will be political will to repeal and replace the Electoral College. That day is not today. There is too much inertia in favor of Democratizing Absolutely Everything as a panacea. I think the Constitution’s original plan for electing the President was much better, so I have only sought to cure its defects… but I think it will be a little while before the People start to see the wisdom in that path. When they do start to come around, my proposal is carefully designed to avoid advantaging either present-day political party over the other, which will make it politically viable where many other proposals are not.

However, when it comes down to an Article V Convention and you’re one of the delegates actually crafting amendments, it may become clear that my proposal doesn’t have enough support to pass. In that case, you may need to settle for second-best.

My rule here is simple: any proposal that increases the amount of individual discretion and deliberation in the Electoral College (or whatever replaces it) is good. Any proposal that further weakens the Electoral College (or abolishes it without replacement) is bad. If it gives voters even more direct power to choose the President (fueling all the “tumults and disorders” Hamilton warned us about), vote it down.

Every part of my proposal is there for a reason, but most provisions are not absolutely essential. The double-majority rule could be tweaked. The eighteen-month rule is good but probably doesn’t break the system if you strip it out. And so on.

If someone proposed simply having the House of Representatives elect the President, I’d seriously consider it. I have some concerns about how that might erode the separation of powers and fuel English-style elections where it’s all about the current party leader’s cult of personality… but, on balance, it seems like it’s probably still better than the insane system we currently have. It certainly wouldn’t vomit up Trump and Biden!

If someone took Section 3 of my proposal (which nullifies all elector bindings) and tried to pass that as a standalone amendment, that would not be a complete solution, but it would be a huge step in the right direction. It would essentially overturn the Supreme Court’s decision in Chiafalo v. Washington. I’d vote for it.

Thanks for Reading

I’m down to hear your counter-proposals / critiques / comments / paeans to my wisdom in the comments.

However, fair warning: if you don’t like this proposal, my second-best plan is sortition!

UPDATE 23 May 2024: I have posted a follow-up to this post, in which I reply at length to some of the comments.

On advice from the comments, in the fourth sentence of section one of the proposed amendment text, I changed the word “governor” to “executive authority.” The sentence originally read: “The elector shall be the current governor of the state (or, if the state has no governor, a holder of the office or member of the body in which the state’s supreme executive authority rests).”

Tag: #SomeConstitutionalAmendments

1. …or, at best, by Twitter, YouTube, and Facebook—our good old domestic home grown algorithmic brainworm machines.

Of course, other countries have popular mass elections for leadership of the executive branch. I don’t understand most other countries well enough to say with confidence that they experience all the same pathologies that we do, and (if not) why not. (I learned my lesson, Tarb! I didn’t address as much of your great comment as I planned to, but at least I learned not to address things when I don’t know what I’m talking about!)

However, I do feel confident enough to say that, if my only choices in an election were Boris Johnson or Jeremy Corbyn… or Justin Trudeau and literally any of these people… then I would start doing some real soul-searching about whether my system was fundamentally busted.

2. The second-place finisher in the Democratic primary in 2016, Bernie Sanders, would have made my point just as well—perhaps even better. So would the second-place finisher in the Republican primary, Sen. Ted Cruz.

The primaries aren’t sending us their best. They’re sending people that have a lot of problems, and they’re bringing those problems with us. They’re bringing drugs. They’re bringing crime. They’re rapists. And some, I assume, are good people. We need a total and complete shutdown of the primaries until our country’s representatives can figure out what the hell is going on.

3. Credit for this term is owed to Zeke.

4. The twelfth, New York, diffused veto power among a state council of revision that included the governor as one member. The thirteenth was Massachusetts. Much of this paragraph is pulled from Akhil Amar’s America’s Constitution: A Biography, pp137-140.

5. …except in New Hampshire and Vermont, where the governor serves only two-year terms. However, neither state has term limits, and every governor for the past six decades has served at least four years in office, with but two exceptions: Vermont Gov. Richard Snelling died in office during his first year, and New Hampshire Gov. Craig Benson, who lost a close race twenty years ago.

6. The occasional revolt still happens. Alaska Governor Frank Murkowski was profoundly unpopular, and lost the gubernatorial primary to the mayor of Wasila, an up-and-comer named Sarah Palin. Wisconsin Governor Scott McCallum took office without being elected after Tommy Thompson went to Washington. He never enjoyed such broad support from the party. He lost the primary to Scott Walker.

In the past 26 years, across all 50 states, a sitting governor has been defeated in the primary a total of (I think) 5 times. (Murkowski-AK, McCallum-WI, Abercrombie-HI, Gibbons-NV, and Holden-MO.) (Gov. Gray Davis of CA was also successfully recalled.) This suggests that gubernatorial power over the party is generally very strong, and usually falls apart only when the governor is devastated by scandal or otherwise becomes broadly unpopular. Though he is not completely unaccountable to the party, he is surely, in the vast majority of circumstances, less accountable to it than any other state official.

7. I don’t like the Twenty-Third Amendment, but it’s settled law now, not worth fighting over.

8. Moreover, by weighting electoral college votes to actual census population, we solve the Democrats’ longstanding (and quite reasonable) complaint that small states currently enjoy disproportionate advantages in the current Electoral College because of weird quirks in how the College assigns electors to each state. (For example, Wyoming gets 3.5 more electors per person than Florida.) In this proposal, that’s fixed.

9. This includes the Mayor of D.C., a Democrat.

10. It is worth recalling that Hillary beat Trump in the 2016 popular vote—but she didn’t have majority support, either! Only 48% of Americans voted for her! 52% voted for somebody else, and it’s very possible that the actual 2016 Condorcet winner was libertarian candidate Gary Johnson or independent Evan McMullin! This is another reason why arguments for a national popular vote based on “democracy” fall short. The national popular vote often fails to produce Condorcet winners.

11. I’m counting all the internal GOP caucus votes, including the caucus roll-call votes. If I counted the postponed House floor votes that never materialized during Steve Scalise’s short candidacy, I could make it an even twenty, but that seems like cheating.

12. “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.”

Congress has indeed provided by law for succession: 3 USC 19 specifies that the Speaker of the House must resign from the House and serve as “acting President” until a true President is elected. If the Speaker can’t, the president pro tem of the Senate must resign and become acting President. After that, it falls to the Secretary of State, Treasury, Defense, and so on down the line of cabinet secretaries until Acting President Laura Roslin is sworn in.

13. Except in 1840, when a rule change allowed a simple majority to suffice.

14. Prior to 1179, Catholic papal elections were formally unanimous. However, the necessary “unanimity” was often achieved at the point of a sword. If you have a free user account at the Internet Archive, you can take turns reading about this by checking out Frederic Baumgartner’s breezy Behind Locked Doors: A History of Papal Elections. (You can borrow the digital copy of the book for one hour if nobody else has it checked out.) It’s shallow and occasionally gives credence to apocrypha, but it narrates each and every papal election, and hoo boy it’s a bummer how many of them end in an armed party storming St. John Lateran’s basilica and consecrating their preferred candidate by force.

15. Pope St. John Paul II changed canon law so that, after thirty-three ballots, the two-thirds rule could be waived by vote of a simple majority. This, of course, eviscerated the super-majority rule. After all, if a papabile has 55% support but is profoundly opposed by the other 45%, all he and his supporters have to do is wait it out! After thirty-three ballots, they can vote to waive the rule and their guy becomes Pope! As a practical matter, John Paul’s revision eliminated the two-thirds requirement, just with extra steps. Alas, he never seemed to realize it.

St. John Paul was very smart about some things and very stupid about others. One of the things JP2 was very stupid about was governance. This is fine for most people—my mom is the first to admit that she is pretty bad at governance, too—but “bad at governance” is an unfortunate trait in the supreme governing authority of the Church Militant.

Pope Benedict XVI partially restored the old rule in 2007, but not fully: a two-thirds majority is once again required, but, after thirty-three ballots of deadlock, the cardinals are now forced to a runoff election between the top two vote-getters. I admit there are parts of this approach that I like, but, by preventing the cardinals from seeking a compromise candidate after a deadlock, I think it risks creating more deadlock than it solves.

16. This was a weirdly intense deadlock, even for a convention with a super-majority rule. The Democratic National Convention had also deadlocked in 1912 (forty-six ballots) and 1920 (forty-four ballots), but going past a hundred ballots was unheard of. In a 1972 Note in The Journal of Southern History entitled “William Gibbs McAdoo and the Democratic National Convention of 1924,” James C. Prude argues that McAdoo stayed in the race long after it was clear he couldn’t achieve two-thirds because he thought he could eventually, through rules chicanery, win with a simple majority. Smith then stayed in simply to block McAdoo, until both agreed to withdraw on the hundredth ballot.

At no point did either McAdoo or Smith actually receive even a simple majority of the votes cast.

17. Nineteenth century election schedules were much less well-defined than ours. In 1845, Congress standardized that Presidential electors must be chosen on the first Tuesday in November, but it took decades for states to move all their other elections to that date. Check out the free-for-all scheduling in the House elections of 1858-59.

18. Todd Zywicki has quite a bit more to say about this general problem in a 1997 article, Beyond the Shell and Husk of History: the History of the Seventeenth Amendment and Its Implications for Current Reform Proposals, especially in the section headed “Bicameralism.”